Quantum Machine Learning for 6G Communication Networks: State-of-the-Art and Vision for the Future

In the paper

“In this regard, a model-free distributed reinforcement learning method for power allocation is proposed [79], in which Channel State Information (CSI) and QoS indicators are exploited to adapt the transmit power.”

What the sources say

• Disputed

  The Abstract and Introduction describe a reinforcement-learning agent at each eNB, but explicitly identify the adapted variable as the α-fair packet-scheduling parameter. The claimed power-allocation method is contrary to the paper’s stated contribution; the Simulation Scenario lists transmit power only as a simulation parameter. No page numbers are available in our copy. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  cites [79] 79. M. I. Tiwana, S. J. Nawaz, A. A. Ikram, and M. I. Tiwana, "Self-organizing networks: A packet scheduling approach for cover- age/capacity optimization in 4G networks using reinforcement learning," Elektronika ir Elektrotechnika, vol. 20, no. 9, pp. 59-64, 2014. sourceExperimental — accuracy not yet measured

In the paper

“In this context, a DL based framework for optimization of downlink beamforming is proposed in [92].”

What the sources say

• Disputed

  The Introduction (page unavailable) says this paper proposes methods for WLAN indoor localization, including handling missing access points and interpolating location estimates. That account of what the paper proposes conflicts with the attribution of a downlink-beamforming optimization framework. (LLM, conf=0.87)

  Checked by a reasoning model — may be wrong.

  cites [92] 92. F. Saleem and S. Wyne, "WLAN-based indoor localization using neural networks," Journal of Electrical Engineering, vol. 67, no. 4, pp. 299-306, 2016. sourceExperimental — accuracy not yet measured

In the paper

“The Genetic Algorithm (GA) at physical layer of communication systems has been used for optimal antenna selection in MIMO systems, power control, and symbol detection in MIMO systems in [80,81], [82], [83,84], respectively.”

What the sources say

• Disputed

  The citation to this paper is attached to the power-control example, not the antenna-selection or symbol-detection examples. The Introduction (p. 1) identifies its approach as deep reinforcement learning for power control, and the Conclusion (p. 11) describes a deep Q-learning power-allocation algorithm—not a genetic algorithm. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  cites [80] 80. N. Y. S and D. Guo, "Deep reinforcement learning for distributed dynamic power allocation in wireless networks," arXiv preprint arXiv:1808.00490, 2018. sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [81] 81. J. Lain, "Joint transmit/receive antenna selection for MIMO systems: A real-valued genetic approach," IEEE Commun. Letters, vol. 15, no. 1, pp. 58-60, Jan. 2011. source
• ?Not checked — no full text of the sourcecites [82] 82. P. D. Karamalis, N. D. Skentos, and A. G. Kanatas, "Selecting array con- figurations for MIMO systems: an evolutionary computation approach," IEEE Trans. Wireless Commun., vol. 3, no. 6, pp. 1994-1998, Nov. 2004. source
• ?Not checked — no full text of the sourcecites [83] 83. B.-K. Lee, H.-W. Chen, and B.-S. Chen, "Power control of cellular radio systems via robust smith prediction filter," IEEE Trans. Wireless Commun., vol. 3, no. 5, pp. 1822-1831, Sep. 2004. source
• ?Not checked — no full text of the sourcecites [84] 84. H. Y.Lu and W. H. Fang, "Joint receive antenna selection and symbol detection for mimo systems: a heterogeneous genetic approach," IEEE Commun. Letters, vol. 13, no. 2, pp. 97-99, Feb. 2009. source

In the paper

“The 5G networks are aimed at providing a 1000-fold increase in the aggregate throughput and a 10-fold increase in the individual link throughput relative to the 4 th Generation (4G) wireless networks [46].”

What the sources say

• Partly supported

  The “Data Rate” section supports a roughly 1000-fold increase in aggregate data rate from 4G to 5G. It does not support the claimed 10-fold increase in individual-link throughput: it describes the 5G edge-rate goal as about 100-fold above 4G, and the same section says per-link rates will increase by about 100-fold. (LLM, conf=0.93)

  cites [46] 46. J. G. Andrews, S. Buzzi, W. Choi, S. V. Hanly, A. Lozano, A. C. K. Soong, and J. C. Zhang, "What will 5G be?" IEEE J. Sel. Areas Com- mun., vol. 32, no. 6, pp. 1065-1082, Jun. 2014. sourceExperimental — accuracy not yet measured

In the paper

“The QC-assisted communications is another new research area, which is envisioned to hold promise for achieving extremely high data rates and link security in future 6G and beyond communications [30].”

What the sources say

• Partly supported

  The introduction (‘The next generation of wireless communications,’ page not provided) says quantum computing is being combined with classical communications to pursue throughput targets. It does not establish that quantum-assisted communications will improve link security or specifically situate that promise in 6G and beyond. (LLM, conf=0.84)

  cites [30] 30. P. Botsinis, D. Alanis, Z. Babar, H. Nguyen, D. Chandra, S. X. Ng, and L. Hanzo, "Quantum algorithms for wireless communications," IEEE Commun. Surveys Tuts., vol. PP, Nov. 2018. sourceExperimental — accuracy not yet measured

In the paper

“Massive Multiple-Input Multiple-Output (M-MIMO) systems and millimeter-Wave (mmWave) spectrum exploiting high spatial resolution and multi-gigahertz bandwidth, respectively, are believed to have an important role in addressing the capacity demands of future communication networks [12,22].”

What the sources say

• ?Could not verify

  The paper’s Abstract and “Moore’s law and superconducting electronics” section discuss quantum machine learning and hybrid computers, not wireless-network capacity. Searches for massive MIMO, millimeter-wave spectrum, spatial resolution, and bandwidth found no passage addressing the attributed claim. (LLM, conf=0.77)

  cites [12] 12. M. G. Kibria, K. Nguyen, G. P. Villardi, O. Zhao, K. Ishizu, and F. Ko- jima, "Big data analytics, machine learning, and artificial intelligence in next-generation wireless networks," IEEE Access, vol. 6, pp. 32 328- 32 338, May 2018. sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page unavailable) says mmWave offers the large bandwidth needed for high-data-rate vehicular communications, supporting that part of the claim. The paper discusses MIMO channel modeling and large antenna arrays, but the passages found do not say that massive MIMO's spatial resolution helps meet future network capacity demands. (LLM, conf=0.85)

  cites [22] 22. F. Jameel, S. Wyne, S. J. Nawaz, and Z. Chang, "Propagation channels for mmWave vehicular communications: State-of-the-art and future research directions," IEEE Wireless Communications, vol. 26, no. 1, pp. 144-150, Feb. 2019. sourceExperimental — accuracy not yet measured

In the paper

“ML can potentially assist big data analytics to realize self-sustaining and proactive wireless networks [12].”

What the sources say

• ?Could not verify

  The Abstract discusses potential uses of quantum machine learning, and Societal Implications discusses computerized agents and web traffic. Neither addresses big-data analytics for self-sustaining, proactive wireless networks; searches for those terms and their paraphrases found no supporting passage in our copy. (LLM, conf=0.75)

  cites [12] 12. M. G. Kibria, K. Nguyen, G. P. Villardi, O. Zhao, K. Ishizu, and F. Ko- jima, "Big data analytics, machine learning, and artificial intelligence in next-generation wireless networks," IEEE Access, vol. 6, pp. 32 328- 32 338, May 2018. sourceExperimental — accuracy not yet measured

In the paper

“The TI allows for a real-time interaction of humans and machines that supports haptic input with audiovisual feedback to control the machine operation in real-time.”

What the sources say

• Supported

  Section IV describes real-time haptic control through the Internet, with audio/visual feedback and, for haptic control, additional touch feedback. Section IX says the Tactile Internet enables real and virtual objects to be steered at real-time interaction speeds. Together, these support the cited description of human–machine control. (LLM, conf=0.90)

  cites [56] 56. M. Simsek, A. Aijaz, M. Dohler, J. Sachs, and G. Fettweis, "5G-enabled tactile internet," IEEE J. Sel. Areas Commun., vol. 34, no. 3, pp. 460- 473, Mar. 2016. source

In the paper

“A bank of ANNs is proposed for symbol decoding in MIMO-OFDM systems in [74].”

What the sources say

• ?Could not verify

  The Introduction describes a lightweight quantum-cryptography scheme for transmitting smart-grid power data, and the cryptographic-algorithm section describes a stream cipher. Searches for a bank of ANNs, symbol decoding, and MIMO-OFDM found no passage addressing the attributed proposal in our copy. (LLM, conf=0.76)

  cites [74] 74. Y. Li, P. Zhang, and R. Huang, "Lightweight quantum encryption for secure transmission of power data in smart grid," IEEE Access, vol. PP, 2019. sourceExperimental — accuracy not yet measured

In the paper

“ANNs are considered as the most notable enablers of DL mechanism [17], in which the understanding of the data's distribution and important features are automatically learned to simulate an effective mapping function.”

What the sources say

• Supported

  Section III (p. 13) calls ANNs enablers of deep learning and says they learn complex input-output relationships and statistical structures from observed data. Section II-C (p. 13) says deep learning uses ANNs to learn the underlying data distribution and extract features needed for a good mapping function. (LLM, conf=0.94)

  cites [17] 17. M. D. A. Zappone, M. D. Renzo, "Wireless networks design in the era of deep learning: Model-based, AI-based, or both?" arXiv preprint arXiv:1902.02647, 2019. source

In the paper

“The scope of GA for intelligent cognitive radio has been encouraged in [87].”

What the sources say

• Partly supported

  The Introduction (p. 1) describes GAs as useful for optimization and applies them to wireless-channel estimation; Section III (p. 4) reports improved performance for the proposed GA-based estimator. Our copy does not discuss intelligent cognitive radio or encourage GA use for that specific purpose (searched: cognitive radio, cognitive). (LLM, conf=0.78)

  cites [87] 87. S. J. Nawaz, M. I. Tiwana, M. N. Patwary, N. M. Khan, M. I. Tiwana, and A. Haseeb, "GA based sensing of sparse multipath channels with superimposed training sequence," Elektronika ir Elektrotechnika, vol. 22, no. 1, pp. 87-91, 2016. sourceExperimental — accuracy not yet measured

In the paper

“The scope of GA for estimation of such channels has been investigated in the literature, see e.g., [84,86] for GA based estimation of sparse channels (UWAC), etc.”

What the sources say

• ?Not checked — no full text of the sourcecites [84] 84. H. Y.Lu and W. H. Fang, "Joint receive antenna selection and symbol detection for mimo systems: a heterogeneous genetic approach," IEEE Commun. Letters, vol. 13, no. 2, pp. 97-99, Feb. 2009. source
• ?Could not verify

  Section 3 (Related work; page unavailable) discusses genetic algorithms for wireless-communications problems, but does not address estimation of sparse underwater acoustic channels. Searches for “UWAC,” “underwater acoustic sparse channel,” and GA-based sparse-channel estimation found no passage addressing that specific application. (LLM, conf=0.72)

  cites [86] 86. S. K. Goudos, "Evolutionary algorithms for wireless communications -a review of the state-of-the-art," in Contemporary Issues in Wireless Communications. InTech, 2014. sourceExperimental — accuracy not yet measured

In the paper

“Deployment of ANNs in communication networks is not a new idea, instead, ANNs have been deployed to perform and assist in various operations of communication systems; e.g., ANN is proposed for symbol decoding for MIMO-OFDM systems in [74].”

What the sources say

• ?Could not verify

  The Introduction and the section on quantum cryptography for smart grids discuss secure power-data communication, not ANN deployment. A search of our copy for ANN-based MIMO-OFDM symbol decoding found no passage addressing the cited example; the available passages therefore cannot verify the attribution. (LLM, conf=0.76)

  cites [74] 74. Y. Li, P. Zhang, and R. Huang, "Lightweight quantum encryption for secure transmission of power data in smart grid," IEEE Access, vol. PP, 2019. sourceExperimental — accuracy not yet measured

In the paper

“Description of radio propagation channels characteristics for molecular communications is not well established in the literature [89]; for such communication scenarios, an ANN based receiver design is proposed in [90].”

What the sources say

• ?Could not verify

  The Introduction discusses millimeter-wave cellular channels, and the Problem Formulation describes neural networks that predict channel features at a base station. Neither passage addresses whether propagation-channel characteristics for molecular communications are established or proposes an ANN-based receiver for that setting. Searches for molecular communications and ANN receiver design found no supporting passage in our copy. (LLM, conf=0.77)

  cites [89] 89. S. Navabi, C. Wang, O. Y. Bursalioglu, and H. Papadopoulos, "Pre- dicting wireless channel features using neural networks," arXiv preprint arXiv:1802.00107, 2018. sourceExperimental — accuracy not yet measured
• Supported

  The citation to this source is attached to the channel-characterization claim, not the ANN receiver proposal, which is cited to [90]. The survey says molecular-communication channels differ from radio-based channels and require new models (Section V, p. 13); it also identifies a need for more realistic, experimentally validated propagation models (Section A, p. 25). (LLM, conf=0.86)

  cites [90] 90. N. Farsad, H. B. Yilmaz, A. Eckford, C. Chae, and W. Guo, "A com- prehensive survey of recent advancements in molecular communication," IEEE Commun. Surveys Tuts., vol. 18, no. 3, pp. 1887-1919, Feb. 2016. source

In the paper

“applications of DL in communication networks have been indicated in the literature, which includes network planning, deployment, resources management, operations control, and maintenance etc.”

What the sources say

• Supported

  The paper has sections on deep learning for network deployment and planning (Section C, p. 4), resource management (Section D, p. 7), and operation and maintenance (Section E, p. 9). Those sections support the citing sentence’s list of application areas. (LLM, conf=0.92)

  cites [17] 17. M. D. A. Zappone, M. D. Renzo, "Wireless networks design in the era of deep learning: Model-based, AI-based, or both?" arXiv preprint arXiv:1902.02647, 2019. source

In the paper

“The DNN for real-time radio resource management at the physical layer of communication networks has been proposed in [101].”

What the sources say

• Supported

  The Introduction (p. 1) says the authors design an algorithm for transmitters to compute power allocation in real time and describes its use of a pretrained neural network. The paper thus proposes a neural-network-based approach to real-time wireless radio resource management; its specific resource-management task is transmit power allocation. (LLM, conf=0.90)

  cites [101] 101. T. Wang, C.-K. Wen, H. Wang, F. Gao, T. Jiang, and S. Jin, "Deep learning for wireless physical layer: Opportunities and challenges," China Communications, vol. 14, no. 11, pp. 92-111, 2017. source

In the paper

“Some of the promising quantum communication protocols to expand the possibility of classical data transmission in quantum-based systems include quantum key distribution (QKD) [126,127], quantum teleportation [128] and dense coding [129].”

What the sources say

• ?Could not verify

  The Summary of Results (p. 5) discusses quantum-measurement simulation and channel coding, not a list of promising communication protocols. Searches for quantum key distribution, teleportation, and dense coding found only a QKD reference in the bibliography (p. 44), with no passage supporting the attributed claim. (LLM, conf=0.76)

  cites [126] 126. M. Berta, J. M. Renes, and M. M. Wilde, "Identifying the information gain of a quantum measurement," IEEE Trans. Inf. Theory, vol. 60, no. 12, pp. 7987-8006, Dec. 2014. sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [127] 127. V. Sharma and S. Banerjee, "Analysis of quantum key distribution based satellite communication," in proc. of Int. Conf. on Computing, Commun. and Networking Technol., Jul. 2018, pp. 1-5. source
• ?Not checked — no full text of the sourcecites [128] 128. K. Inoue, "Quantum key distribution technologies," IEEE J. Sel. Topics Quantum Electron., vol. 12, no. 4, pp. 888-896, Jul. 2006. source
• Supported

  The cited paper is the quantum-teleportation protocol identified by citation [128]. Its discussion of sending an unknown quantum state through classical and EPR channels, and of teleportation's potential usefulness (passages 1 and 8), supports that identification. QKD and dense coding have separate citations in the citing sentence. (LLM, conf=0.91)

  cites [129] 129. C. H. Bennett, G. Brassard, C. Crépeau, R. Jozsa, A. Peres, and W. K. Wootters, "Teleporting an unknown quantum state via dual classical and einstein-podolsky-rosen channel," Physical Review Letters, vol. 70, no. 13, pp. 1895-1899, Mar. 1993. source

In the paper

“The quantumassisted communications can enhance various aspects of the existing classical communication networks including channel estimation, optimal Multi-User Detection (MUD), the design of optimal precoding matrix and the optimal routing by employing the quantum algorithms [30].”

What the sources say

• Supported

  In “A. Why Quantum Computing?”, the paper says quantum algorithms may improve classical communication processes and explicitly lists multi-user detection, channel estimation, downlink precoding-matrix selection, and route finding. The claim preserves that qualified, potential-benefit framing. (LLM, conf=0.95)

  cites [30] 30. P. Botsinis, D. Alanis, Z. Babar, H. Nguyen, D. Chandra, S. X. Ng, and L. Hanzo, "Quantum algorithms for wireless communications," IEEE Commun. Surveys Tuts., vol. PP, Nov. 2018. source

In the paper

“Such dynamic sharing of spectrum aims at enhancing the spectrum utilization efficiency through the provision of access of under-utilized spectrum to the secondary network users [111].”

What the sources say

• Supported

  The Introduction (p. 1) describes spectrum as underutilized and says sharing it with secondary users is motivated by the goal of improving spectrum efficiency. It also states that secondary-network users may access spectrum owned by licensed users, provided they do not cause harmful interference. (LLM, conf=0.94)

  cites [111] 111. X. Li, J. Fang, W. Cheng, H. Duan, Z. Chen, and H. Li, "Intelligent power control for spectrum sharing in cognitive radios: A deep reinforcement learning approach," IEEE Access, vol. 6, pp. 25 463-25 473, Apr. 2018. source

In the paper

“Cognitive radio techniques enables a radio system to sense, learn, and adapt based on the context of the surrounding environment [87].”

What the sources say

• ?Could not verify

  The Introduction (p. 1) discusses wireless-channel estimation and genetic algorithms, not cognitive radio techniques. Searches for cognitive radio and for sensing, learning, and adapting to environmental context found no passage addressing the attributed claim. (LLM, conf=0.76)

  cites [87] 87. S. J. Nawaz, M. I. Tiwana, M. N. Patwary, N. M. Khan, M. I. Tiwana, and A. Haseeb, "GA based sensing of sparse multipath channels with superimposed training sequence," Elektronika ir Elektrotechnika, vol. 22, no. 1, pp. 87-91, 2016. sourceExperimental — accuracy not yet measured

In the paper

“In [109], supervised and unsupervised ML-based cooperative spectrum sensing algorithms for cognitive radio networks have been proposed.”

What the sources say

• Partly supported

  The survey describes cooperative spectrum sensing in “A. Spectrum Awareness Techniques” (p. 5) and a learning approach used for feature detection in “A. Cyclostationary Based Detector” (p. 11). Neither passage identifies supervised and unsupervised ML-based cooperative sensing algorithms, and the cited work is a survey rather than a paper proposing them. (LLM, conf=0.82)

  cites [109] 109. S. K. Sharma, T. E. Bogale, S. Chatzinotas, B. Ottersten, L. B. Le, and X. Wang, "Cognitive radio techniques under practical imperfections: A survey," IEEE Commun. Surveys Tuts., vol. 17, no. 4, pp. 1858-1884, Jul. 2015. sourceExperimental — accuracy not yet measured

In the paper

“Quantum-proof randomness extractors can be considered as an important building block for implementing classical and quantum cryptography in security applications [160].”

What the sources say

• Supported

  The Introduction (p. 1) calls extractors an essential tool in cryptography and describes applications of quantum-secure privacy amplification in quantum cryptography. The Introduction (p. 2) also studies extractors for cryptographic settings involving quantum adversaries. Together, these passages support the broad building-block claim. (LLM, conf=0.91)

  cites [160] 160. R. T. Konig and B. M. Terhal, "The bounded-storage model in the presence of a quantum adversary," IEEE Trans. Inf. Theory, vol. 54, no. 2, pp. 749-762, Feb. 2008. source

In the paper

“For example, the achievable rate for the transmission of classical information over a noisy quantum channel is given by the Holevo-Schumacher-Westmorel (HSW) coding theorem [136], which generalizes the Shannon's theorem in quantum settings.”

What the sources say

• ?Could not verify

  The Experimental results (p. 4) discuss quantum annealing and success probabilities, while the Conclusions (p. 5) discuss possible quantum speedup. Neither addresses the achievable rate for transmitting classical information over a noisy quantum channel. Searches for the HSW coding theorem, channel capacity, and Shannon's theorem found no relevant passage in our copy. (LLM, conf=0.77)

  cites [136] 136. S. Boixo, T. F. Rønnow, S. V. Isakov, Z. Wang, D. Wecker, D. A. Lidar, J. M. Martinis, and M. Troyer, "Evidence for quantum annealing with more than one hundred qubits," Nature Physics, vol. 10, no. 3, p. 218â Ȃ Ş224, Feb. 2014. sourceExperimental — accuracy not yet measured

In the paper

“For example, authors in [126] discussed and analyzed the applicability of QKD protocols in quantumassisted SatCom systems in order to perform secure communication between ground stations and the satellite.”

What the sources say

• ?Could not verify

  The Summary of Results (p. 5) describes measurement simulation and state-splitting protocols, while Extensions and Applications (pp. 30–31) discusses information reconciliation and quantum side information. Searches for satellite, SatCom, QKD, and quantum key distribution found no passage addressing QKD between ground stations and a satellite. (LLM, conf=0.77)

  cites [126] 126. M. Berta, J. M. Renes, and M. M. Wilde, "Identifying the information gain of a quantum measurement," IEEE Trans. Inf. Theory, vol. 60, no. 12, pp. 7987-8006, Dec. 2014. sourceExperimental — accuracy not yet measured

In the paper

“and IBM has a recent plan of finalizing a gate-based quantum computer with 50 qubits by 2020 [30].”

What the sources say

• Supported

  The section “A. Why Quantum Computing?” states that IBM had announced plans to deliver a 50-qubit gate-based quantum computer by 2020. That directly supports the attributed claim as a report of IBM’s plan, not a claim that the computer was completed. (LLM, conf=0.95)

  cites [30] 30. P. Botsinis, D. Alanis, Z. Babar, H. Nguyen, D. Chandra, S. X. Ng, and L. Hanzo, "Quantum algorithms for wireless communications," IEEE Commun. Surveys Tuts., vol. PP, Nov. 2018. source

In the paper

“QML is observed to exponentially speed-up the optimization problems involving quadratic objective functions subject to equality constraints and involving penalty functions subject to inequality constraints [41].”

What the sources say

• Partly supported

  The qBLAS-based optimization section (p. 8) describes an exponential algorithmic speedup for quadratic objectives subject to equality constraints, but only when the matrices are sparse or low rank. It also says inequality constraints are often handled with penalty functions and describes an exponentially faster method for polynomial optimization; it does not establish that all such constrained problems are sped up. The Perspectives section (p. 14) treats practical performance and benchmarking as open questions, so “observed” overstates the evidence. (LLM, conf=0.88)

  cites [41] 41. J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, "Quantum machine learning," Nature, vol. 549, no. 7671, pp. 195-202, Sep. 2017. sourceExperimental — accuracy not yet measured

In the paper

“To this end, the concept of grid computing may further assist in realizing the ambitions by conceding the coordinated resource sharing and access to dynamic multi-institutional virtual organizations [197].”

What the sources say

• ?Could not verify

  The Introduction (pp. 1–2) and Conclusion (p. 11) discuss a space-efficient quantum computer simulator and possible FPGA implementation. Searches for grid computing, coordinated resource sharing, and multi-institutional virtual organizations found no passage addressing the attributed claim. (LLM, conf=0.77)

  cites [197] 197. M. P. Frank, L. Oniciuc, U. H. M.-Baese, and I. Chiorescu, "A space- efficient quantum computer simulator suitable for high-speed FPGA implementation," in Quantum Information and Computation VII, vol. 7342. International Society for Optics and Photonics, 2009, p. 734203. sourceExperimental — accuracy not yet measured

In the paper

“Also, ML for optimal user clustering and power allocation in mmWave NOMA has been proposed in [231].”

What the sources say

• Partly supported

  The Introduction and Section III describe an online-learning method for detecting signals in a clustered NOMA uplink. They do not describe an ML method for optimizing user clustering or power allocation, or identify the setting as mmWave NOMA. The citation supports the ML-for-NOMA aspect, but not the proposed optimization attributed to it. (LLM, conf=0.86)

  cites [231] 231. D. A. Awan, R. L. G. Cavalcante, M. Yukawa, and S. Stanczak, "Detec- tion for 5G-NOMA: An online adaptive machine learning approach," in proc. of IEEE Int. Conf. on Commun., May 2018, pp. 1-6. sourceExperimental — accuracy not yet measured

In the paper

“Online adaptive ML approach for detection in NOMA in the context of 5G networks has been proposed in [230].”

What the sources say

• ?Could not verify

  The paper’s discussion of NOMA in 5G describes dynamic spectrum allocation, and its resource-allocation formulation uses successive interference cancellation. No passage found in our copy describes a proposed online adaptive machine-learning approach for NOMA detection (searched for that phrase and for machine-learning detection). Page numbers are unavailable in the indexed copy. (LLM, conf=0.72)

  cites [230] 230. R. Abozariba, M. K. Naeem, M. Patwary, M. Seyedebrahimi, and P. Bull, "NOMA based resource allocation and mobility enhancement framework for IoT in next generation cellular networks," IEEE Access, vol. PP, pp. 1-15, Jan. 2019. sourceExperimental — accuracy not yet measured

In the paper

“For such an FDD M-MIMO system, a dictionary learning-based channel estimation method has been proposed in [225].”

What the sources say

• ?Could not verify

  The Introduction describes end-to-end transmitter and receiver learning, and the Transmitter and Receiver Architectures section describes estimating a Rayleigh block-fading channel response. Searches for dictionary learning and FDD channel estimation found no passage proposing the claimed method for FDD massive-MIMO systems; the channel-response estimate alone does not establish that claim. (LLM, conf=0.76)

  cites [225] 225. F. A. Aoudia and J. Hoydis, "End-to-end learning of communications systems without a channel model," arXiv preprint arXiv:1804.02276, 2018. sourceExperimental — accuracy not yet measured

In the paper

“In addition, NOMA-based MEC network has also received considerable attention in the recent years, see e.g., [233,234].”

What the sources say

• ?Not checked — no full text of the sourcecites [233] 233. G. Gui, H. Huang, Y. Song, and H. Sari, "Deep learning for an effective nonorthogonal multiple access scheme," IEEE Trans. Veh. Technol., vol. 67, no. 9, pp. 8440-8450, Sep. 2018. source
• Supported

  The Introduction (p. 1) describes several recent studies combining NOMA and MEC, and the paper itself investigates a NOMA-based MEC network. This supports the broad statement that the topic had received considerable attention. (LLM, conf=0.90)

  cites [234] 234. Y. Ye, G. Lu, R. Q. Hu, and L. Shi, "On the performance and op- timization for MEC networks using uplink NOMA," arXiv preprint arXiv:1901.04149, 2019. source

In the paper

“The foregoing discussions suggest that the 6G communications will leverage robust learning capability at different network layers to perform diverse tasks such as network management, radio resource allocation, data security, and manipulation of smart surfaces to name a few.”

In the paper

“Both the reliability aspect, with packet error rates ≤ 10 -5 , and end-to-end latencies ∼1ms aim at supporting new use cases such as factory automation, autonomous driving, e-health, building automation, and smart cities, to name a few [50].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [50] 50. International Telecommunication Union, "IMT Visionâ ȂŤ Framework and Overall Objectives of the Future Development of IMT for 2020 and Beyond," IMT, Tech. Rep. document M Series 2083, Sep. 2015.

In the paper

“For example, the RAN congestion in ultra-dense IoT networks may arise due to various reasons including the massive number of short-packet transmissions, huge signalling overhead per data packet, and very dynamic and sporadic nature of device transmissions [70].”

What the sources say

• ?Not checked — no full text of the sourcecites [70] 70. X. Artiga, A. Perez-Neira, J. Baranda, E. Lagunas, S. Chatzinotas, R. Zetik, P. Gorski, K. Ntougias, D. Perez, and G. Ziaragkas, "Shared access satellite-terrestrial reconfigurable backhaul network enabled by smart antennas at mmwave band," IEEE Network, vol. 32, no. 5, pp. 46- 53, Sep. 2018. source

In the paper

“Moreover, ML-based proactive content caching at the edge node can also be exploited to avoid backhaul congestion and further reduce the service latency [66,67].”

What the sources say

• ?Not checked — no full text of the sourcecites [66] 66. T. D. Ponnimbaduge Perera, D. N. K. Jayakody, S. K. Sharma, S. Chatzinotas, and J. Li, "Simultaneous wireless information and power transfer (SWIPT): Recent advances and future challenges," IEEE Com- mun. Surveys Tuts., vol. 20, no. 1, pp. 264-302, Feb. 2018. source
• ?Not checked — no full text of the sourcecites [67] 67. T. Hou, G. Feng, S. Qin, and W. Jiang, "Proactive content caching by exploiting transfer learning for mobile edge computing," in proc. of IEEE Global Commun. Conf., Dec. 2017, pp. 1-6. source

In the paper

“For example, ML techniques can be significantly advantageous in addressing the access congestion problem in emerging ultra-dense IoT networks [14].”

What the sources say

• ?Not checked — no full text of the sourcecites [14] 14. S. K. Sharma and X. Wang, "Collaborative distributed Q-learning for RACH congestion minimization in cellular IoT networks," IEEE Com- munications Letters, pp. 1-1, 2019. source

In the paper

“To the authors' best knowledge, this work is the first to explicitly investigate the deployment of the QML framework for future communication networks.”

In the paper

“The mMTC service aims at providing internet connectivity to such devices.”

In the paper

“However, DL methods lack an efficient mechanism for prior evaluation of the best choice of training algorithm, size and structure of DNN, and parameters setting that suits the model/problem under consideration.”

In the paper

“Both the reliability aspect, with packet error rates ≤ 10 -5 , and end-to-end latencies ∼1ms aim at supporting new use cases such as factory automation, autonomous driving, e-health, building automation, and smart cities, to name a few [50].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [50] 50. International Telecommunication Union, "IMT Visionâ ȂŤ Framework and Overall Objectives of the Future Development of IMT for 2020 and Beyond," IMT, Tech. Rep. document M Series 2083, Sep. 2015.

In the paper

“One of the potential applications of employing DL methods in such mmWave M-MIMO systems is the estimation of radio channel quality [23], which is essential for the design of transmission techniques such as beamforming.”

What the sources say

• ?Not checked — no full text of the sourcecites [23] 23. H. He, C.-K. Wen, S. Jin, and G. Y. Li, "Deep learning-based channel estimation for beamspace mmWave massive MIMO systems," arXiv preprint arXiv:1802.01290, 2018. source

In the paper

“Furthermore, ML has a strong potential to replace the conventional mathematical mode-based algorithmic solutions, given the availability of adequate data and computational power.”

In the paper

“As an example, DL capabilities can be utilized for channel estimation and symbol detection in Orthogonal Frequency-Division Multiplexing (OFDM) systems [20].”

What the sources say

• ?Not checked — no full text of the sourcecites [20] 20. H. Ye, G. Y. Li, and B. Juang, "Power of deep learning for channel esti- mation and signal detection in OFDM systems," IEEE Wireless Commun. Lett., vol. 7, no. 1, pp. 114-117, Feb. 2018. source

In the paper

“Following the evolution trends of previous mobile network generations, the bit/s throughput targets in 6G networks are expected to increase by an order of magnitude relative to those of 5G.”

In the paper

“For example, a Deep Reinforcement Learning (DRL) based decentralized resource allocation mechanism can be utilized to support highly dynamic communication applications [21].”

What the sources say

• ?Not checked — no full text of the sourcecites [21] 21. H. Ye and G. Y. Li, "Deep reinforcement learning for resource allocation in V2V communications," in IEEE Int. Conf. on Commun., May 2018, pp. 1-6. source

In the paper

“These UAV BSs can not only offload the data-traffic of the static BSs but they can also relocate dynamically to provide a more favorable propagation channel to the edge users.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [59] 59. S. J. Nawaz et al.: Quantum ML for 6G Communication Networks: Vision for the Future full-duplex technology: Recent advances and research challenges," IEEE Commun. Surveys Tuts., vol. 20, no. 1, pp. 674-707, Feb. 2018.

In the paper

“higher mobility causes higher Doppler spread which further causes reduction in the coherence-time and this eventually leads to fast variability in the channel statistics.”

In the paper

“Modelbased learning usually optimizes the performance indexes through available objectives functions with high computational efficiency.”

In the paper

“A promising application of reinforcement learning at the physical layer of communication networks is power control and optimization.”

In the paper

“The integration of DL capability with the smart city infrastructure can help in effective utilization of the available big data for accomplishing the dream of the cognitive smart world of the future [27].”

What the sources say

• ?Not checked — no full text of the sourcecites [27] 27. M. Mohammadi and A. Al-Fuqaha, "Enabling cognitive smart cities using big data and machine learning: Approaches and challenges," IEEE Commun. Mag., vol. 56, no. 2, pp. 94-101, Feb. 2018. source

In the paper

“In addition, DL for radio resource management in 5G networks is also suggested in [94], and a users location-aware DL method for run-time optimal users association for the maximization of sum-rate in M-MIMO based land mobile radio cellular networks is proposed in [95].”

What the sources say

• ?Not checked — no full text of the sourcecites [94] 94. Z. Qin, H. Ye, G. Y. Li, and B.-H. F. Juang, "Deep learning in physical layer communications," proc. of IEEE Veh. Technol. Conf. (Fall-19), vol. preprint arXiv:1807.11713, 2019. source
• ?Not checked — no full text of the sourcecites [95] 95. F. D. Calabrese, L. Wang, E. Ghadimi, G. Peters, L. Hanzo, and P. Soldati, "Learning radio resource management in rans: Framework, opportunities, and challenges," IEEE Commun. Mag., vol. 56, no. 9, pp. 138-145, Sep. 2018. source

In the paper

“A typical application of off-line (batch) and online learning in communication systems can be intelligent caching and channel tracking, respectively.”

In the paper

“The selection of encoding/precoding schemes for performance optimization is another potential application of unsupervised learning.”

In the paper

“The generalization of an ANN's operations, for efficiently dealing with every new unseen input data sample, needs its training over a large amount of data (i.e., big data).”

In the paper

“Subsequently, at the higher layers, there are various potential applications of unsupervised and semi-supervised learning for grouping/pairing/clustering of nodes/points for optimal allocation of network/radio resources.”

In the paper

“Recently, an ANN-based method for predicting channel features for large-scale multi-antenna BSs is proposed in [88].”

What the sources say

• ?Not checked — no full text of the sourcecites [88] 88. T. W. Rondeau, B. Le, C. J. Rieser, and C. W. Bostian, "Cognitive radios with genetic algorithms: Intelligent control of software defined radios," in proc. of SDR Forum Technical Conf., vol. 100. Citeseer, 2004, pp. 3-8. source

In the paper

“Unsupervised learning can potentially be applied for performing the wide range of tasks related to points clustering, features extraction, features classification, distribution estimation, and distribution specific samples generation.”

In the paper

“For example, the revisited learning algorithms for exploiting the numerous amount of available processing cores in advanced Graphics Processing Units (GPUs) has demonstrated a remarkable performance gain.”

In the paper

“At the physical layer in highly dynamic scenarios of vehicular communications, where less coherence-time limits the available time and data for supervising the learning of channel equalizer; semi-supervised and unsupervised learning can make their way for assisting the channel equalization and tracking operations.”

In the paper

“Another application of supervised learning at the physical layer for downlink communications can be optimal power allocation and interference cancellation.”

In the paper

“The size of an ANN (number of neurons and hidden layers) and the amount of available data determine its performance and requisite computational power, as illustrated in Fig. 2.”

In the paper

“With 5G reaching its limits in the next decade or so, the design goals for its successor have already begun to be explored in the literature.”

In the paper

“On-line training is a suitable solution for such streaming data applications.”

In the paper

“A typical application of model-based and samples-based learning can be symbol decoding and contents demand prediction, respectively.”

In the paper

“The 5G wireless networks have recently started to be deployed in some parts of the world but the goal of a fully intelligent network furnishing everything as a service and rendering a completely immersive user experience remains elusive.”

In the paper

“To this end, ML and QC methods can play a significant enabling role.”

In the paper

“Furthermore, the advanced Tensor Processing Units (TPUs) have demonstrated a tremendous parallel processing potential with manifold speed-ups and power efficiency in executing ML algorithms.”

In the paper

“at the physical layer, the received noisy data symbols can be used to train a system by clustering the sample points in the decision space for generating effective nonlinear decision boundaries for mapping of the symbols according to the constellation maps.”

In the paper

“An agent can be associated with each serving station in a cellular network to assist in learning the optimal scheduling parameters to enhance the network Quality-of-Service (QoS) [78].”

What the sources say

• ?Not checked — no full text of the sourcecites [78] 78. M. Chen, W. Saad, C. Yin, and M. Debbah, "Echo state networks for proactive caching in cloud-based radio access networks with mobile users," IEEE Trans. Wireless Commun., vol. 16, no. 6, pp. 3520-3535, Jun. 2017. source

In the paper

“The recent advances in parallel computing capabilities and distributed learning methods have enabled the deployment of data-driven DL approaches to complement the conventional model-based approaches.”

In the paper

“Also, the available state-ofart in ML is isolated in terms of ML techniques as well as their operations across different layers of the protocol stack of communication networks.”

In the paper

“As anticipated in Fig. 2, big data-driven DL can enable the deployment of ANNs in learning complex (high dimensions, a high number of features, a high number of classes, etc) statistical structures and input-output operations of the future communication networks.”

In the paper

“These advances have demonstrated a profound impact of DL-based solutions across various multidimensional signal processing applications, e.g., medical image processing, natural languages processing, and wireless communications, to name a few.”

In the paper

“On the other hand, the pure data samplesbased learning exploits all the available data samples to interpolate and/or extrapolate the samples, with high memory and time requirements.”

In the paper

“The training and processing of data through conventional ML algorithms, executed on conventional Central Processing Units (CPUs) with a limited number of cores, has a limitation of large processing delays.”

In the paper

“Given the increasing complexity of the future communication networks, ANNs have a wide scope of its deployment in performing various diverse tasks, e.g., planning, optimization, estimation, tracking, controlling, and maintaining tasks, etc.”

In the paper

“Also, ANNs assisted indoor localization method exploiting two variants of the fingerprinting approaches is proposed in [91].”

What the sources say

• ?Not checked — no full text of the sourcecites [91] 91. X. Qian and M. D. Renzo, "Receiver design in molecular communica- tions: An approach based on artificial neural networks," in Int. Symp. on Wireless Commun. Systems. IEEE, 2018, pp. 1-5. source

In the paper

“This makes the estimation and tracking of such channels a challenging task, where any error in channel estimate can significantly affect the symbol detection performance.”

In the paper

“For such fast timevarying channels, a hybrid learning method is proposed in [75], to assist in estimation and tracking of the channels.”

What the sources say

• ?Not checked — no full text of the sourcecites [75] 75. S. J. Nawaz, S. Mohsin, and A. A. Ikaram, "Neural network based MIMO-OFDM channel equalizer using comb-type pilot arrangement," in proc. of IEEE Int. Conf. on Future Computer and Commun., Apr. 2009. source

In the paper

“Another potential scenario, where supervised ML can play an important role, is in determining the users association with BSs on the basis of contents/media demand.”

In the paper

“ML has a popular application in intelligent caching, which can help in reduc-ing the latency in satellite links.”

In the paper

“However, the major limitation is the available computational power to deal with big data and deep structured ANNs.”

In the paper

“The QC has a recognized potential in conveniently manipulating such large-dimensional data vectors in large tensor product spaces.”

In the paper

“Another popular application of DNNs at the physical layer of communication networks is power control and optimization [102].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [102] 102. H. Sun, X. Chen, Q. Shi, M. Hong, X. Fu, and N. D. Sidiropoulos, "Learning to optimize: Training deep neural networks for wireless re- source management," in proc. of IEEE Int. Workshop on Signal Process- ing Advances in Wireless Commun., Jul. 2017, pp. 1-6.

In the paper

“This learning technique offers the advantage of facilitating the learning process by reducing the amount of required data, and relaxes the condition on the training data to be independently and identically distributed (i.i.d.) with the validation data.”

In the paper

“Although the capacity of classical channels has been well understood within the framework of classical information theory, the capacity of quantum channels is not completely understood and various measures are available in the literature.”

In the paper

“An example of such delay critical applications at the physical layer of communication networks is optimal beamforming; where the latency induced by conventional iterative methods makes it outdated for the future networks.”

In the paper

“For enabling the quantum communications, the information signal can be encoded in different ways such as by modulating the photons' polarization, which is usually detected by utilizing single-photon detectors, and the photons' phase, usually measured by the homodyne detection [120].”

What the sources say

• ?Not checked — no full text of the sourcecites [120] 120. J. D. Franson, "Quantum communications," in proc. of Technical Digest. Summaries of papers presented at the Quantum Electronics and Laser Science Conf. (Post conf. Technical Digest), May 2001, pp. 171-. source

In the paper

“In summary, quantum principles provide significant benefits to the communication networks including enhanced channel capacity, the ability to transmit an unknown quantum state, i.e, quantum teleportation, and to deliver secure information, i.e., quantum cryptography by utilizing a number of advanced communication protocols which will not be possible with the classical techniques [119].”

What the sources say

• ?Not checked — no full text of the sourcecites [119] 119. S. Imre, "Quantum communications: Explained for communication engi- neers," IEEE Commun. Mag., vol. 51, no. 8, pp. 28-35, Aug. 2013. source

In the paper

“In this regard, authors in [125] have shown that shared randomness is available if necessary, quantum measurements can be asymptotically represented by the amount of classical communication equivalent to the quantum notion of the mutual information of the measurement.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [125] 125. I. Devetak, "The private classical capacity and quantum capacity of a quantum channel," IEEE Trans. Inf. Theory, vol. 51, no. 1, pp. 44-55, Jan. 2005.

In the paper

“QC and ML can create close synergies with each other towards providing their joint benefits in communication systems.”

In the paper

“A DNN based auto-encoder for jointly optimizing Bit-Error-Rate (BER) and Peak-to-Average Power Ratio (PAPR) in OFDM systems is proposed in [97].”

What the sources say

• ?Not checked — no full text of the sourcecites [97] 97. N. C. Luong, D. T. Hoang, S. Gong, D. Niyato, P. Wang, Y. C. Liang, and D. I. Kim, "Applications of deep reinforcement learning in commu- nications and networking: A survey," arXiv preprint arXiv:1810.07862, 2018. source

In the paper

“Deep transfer learning is another new direction of research to reduce the dependence of learning from the required large amount of data.”

In the paper

“Regarding the quantum information sources, a single photon source can be considered as an ideal source for generating quantum information and this can generate pulses having the mean number of pulses equal to one and zero variance [120].”

What the sources say

• ?Not checked — no full text of the sourcecites [120] 120. J. D. Franson, "Quantum communications," in proc. of Technical Digest. Summaries of papers presented at the Quantum Electronics and Laser Science Conf. (Post conf. Technical Digest), May 2001, pp. 171-. source

In the paper

“Despite ANNs being promising, the limitation of required computational power for the training of an ANN creates challenges for their practical applications.”

In the paper

“For the first approach, the polarization is not preserved while transmitting quantum signals via optical fibers and it is necessary to have a non-intrusive polarization control to preserve the information being transmitted in the quantum domain.”

In the paper

“However, the realization of such an ideal photon source in practical quantum communications is very challenging since it occupies very large space and requires trained technicians.”

In the paper

“Furthermore, spread spectrum based multiple access techniques, which can send the photons of multiple users via the medium (optical or free-space) by sharing the frequency band, time window and the route, seem promising in the context of quantum communications [130].”

What the sources say

• ?Not checked — no full text of the sourcecites [130] 130. C. H. Bennett and S. J. Wiesner, "Communication via one-and twoparti- cle operators on einstein-podolsky-rosen states," Phys. Rev. Lett., vol. 69, no. 20, p. 2881, Nov. 1992. source

In the paper

“Out of these, fainted lasers is mostly used in quantum cryptographic key distribution systems while the four-wave mixing process is useful for optical processing devices such as parametric amplifiers and wavelength converters.”

In the paper

“The first challenge is regarding the construction of network entities with quantum Internet which requires quantum switches/routers and repeaters, which becomes difficult due to no-cloning theorem 122].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [122] 122. L. Gyongyosi, S. Imre, and H. V. Nguyen, "A survey on quantum channel capacities," IEEE Commun. Surveys Tuts., vol. 20, no. 2, pp. 1149-1205, Jan. 2018.

In the paper

“One important advantage of utilizing quantum domain in communications is high degrees of freedom.”

In the paper

“where the highlighted applications are: dynamic spectrum access, joint user association, data caching and offloading, security, connection preservation, traffic routing, resource sharing, and power control.”

What the sources say

• ?Not checked — no full text of the sourcecites [96] 96. A. Zappone, L. Sanguinetti, and M. Debbah, "User association and load balancing for massive MIMO through deep learning," arXiv preprint arXiv:1812.06905, 2018. source

In the paper

“the recent advances in Graphical processing units (GPUs) have provided the initial breakthrough by speeding up the training process through their ability to simultaneously perform multiple weight calculations operations.”

In the paper

“Furthermore, DNN based architectures for symbol detection in MIMO systems have been proposed in [98], and a DNN-based localization method exploiting fingerprintbased and channel measurements for M-MIMO systems has been proposed in [99].”

What the sources say

• ?Not checked — no full text of the sourcecites [98] 98. M. Kim, W. Lee, and D. Cho, "A novel PAPR reduction scheme for OFDM system based on deep learning," IEEE Commun. Letters, vol. 22, no. 3, pp. 510-513, Mar. 2018. source
• ?Not checked — no full text of the sourcecites [99] 99. N. Samuel, T. Diskin, and A. Wiesel, "Learning to detect," arXiv preprint arXiv:1805.07631, 2018. source

In the paper

“In comparison to the classical binary based communications systems, quantum communications has the great potential to provide absolute randomness and security, to carry much more information and to significantly enhance the transmission quality.”

In the paper

“A deep reinforcement learning based power control method for spectrum sharing in cognitive radios has been proposed in [110].”

What the sources say

• ?Not checked — no full text of the sourcecites [110] 110. K. M. Thilina, K. W. Choi, N. Saquib, and E. Hossain, "Machine learning techniques for cooperative spectrum sensing in cognitive radio networks," IEEE J. Sel. Areas Commun., vol. 31, no. 11, pp. 2209-2221, Nov. 2013. source

In the paper

“A detector design by unfolding iterative calculations into ANN layers for MIMO decode and forward relay channels has been proposed in [106].”

What the sources say

• ?Not checked — no full text of the sourcecites [106] 106. J. R. Hershey, J. L. Roux, and F. Weninger, "Deep unfolding: Model-based inspiration of novel deep architectures," arXiv preprint arXiv:1409.2574, 2014. source

In the paper

“Furthermore, quantum-based techniques are able to execute the tasks much faster and beyond the capability of the classical systems [121].”

What the sources say

• ?Not checked — no full text of the sourcecites [121] 121. A. N. Pinto, N. A. Silva, N. J. Muga, Ã. J. Almeida, and D. F. Pereira, "Quantum communications: An engineering approach," in proc. of Int. Conf. on Transparent Optical Networks, Jul. 2017, pp. 1-4. source

In the paper

“The recent advancements in DNNs has enabled its deployment in even delay critical applications; which is achieved through offline training the DNNs and then performing the online operations (tracking, optimization, etc).”

In the paper

“Furthermore, quantum channel/error correction coding is of significant importance for the practical design of quantumassisted communication protocols to approach closer to the theoretical achievable capacity.”

In the paper

“DL can be seen as an end-to-end solution for replacing the sequential blocks based processing methods for estimation and decoding of information at the receivers.”

In the paper

“While the second approach based on the photons' phase does not require the polarization control but an optical carrier (either propagated along with the quantum signal or generated locally at the receiving side) is needed in order to retrieve the phase information.”

In the paper

“Also, it is envisioned that the combination of QC and ML features together in the framework of QML can produce and recognize the statistical data patterns which classical computers and classical ML are unable to perform effectively.”

In the paper

“In contrast to the previous quantumlike models which mainly utilize several fitting parameters for the construction of some models to describe particular effects of a use-case, the QDT model proposed in [150] is considered as a generic theory applicable to any type of decision making, and its mathematical structure is common to both the decision theory and quantum measurements.”

What the sources say

• ?Not checked — no full text of the sourcecites [150] 150. M. Ashtiani and M. A. Azgomi, "A survey of quantum-like approaches to decision making and cognition," IEEE Trans. Inf. Theory, vol. 75, pp. 49-80, May 2015. source

In the paper

“The recent success of quantumassisted and data-driven learning methods in communication networks (discussed in previous sections) has a clear motivation to consider these as enablers of future heterogeneous networks.”

In the paper

“One of the crucial issues for the application of quantum technology in communications related applications is the presence of harmful quantum perturbations, whose harmful effects can be mitigated by employing Quantum Error Correction Codes (QECCs) [141].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [141] 141. M. Hsieh and M. M. Wilde, "Trading classical communication, quantum communication, and entanglement in quantum shannon theory," IEEE Trans. Inf. Theory, vol. 56, no. 9, pp. 4705-4730, Sep. 2010.

In the paper

“For the realization of such consciousness and responsiveness in massively connected heterogeneous networks of future, very robust processing and learning capabilities will be required.”

In the paper

“Long-distance quantum communication can be suitably realised through the physical platform of photons, where an open research problem is the loss of photons in quantum channels.”

In the paper

“Development of transducers to photonic states can help in traveling long distances with minimal decoherence.”

In the paper

“Recently, Google has announced a 72 qubit superconducting quantum computer [195].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [195] 195. K. Klemic and J. Zeigler, Eds., Future Directions of Quantum Informa- tion Processing: A Workshop on the Emerging Science and Technology of Quantum Computation, Communication, and Measurement. Basic Research Innovation and Collaboration Center in Arlington, VA: Virginia Tech Applied Research Corporation (VT-ARC), Aug. 2016.

In the paper

“This concept helps in reducing the delay and power consumption in data routing and delivery, and it offers a significant performance improvement for all type of users; e.g., the smartphones designed for previous generations also equally benefit from this.”

In the paper

“In this regard, authors in [35] showed the possibilities of utilizing QSA for reducing the computational complexity of mm-Wave based and VLC-based localization algorithms while achieving the same performance as that of a full search.”

What the sources say

• ?Not checked — no full text of the sourcecites [35] 35. P. Botsinis, D. Alanis, S. Feng, Z. Babar, H. V. Nguyen, D. Chandra, S. X. Ng, R. Zhang, and L. Hanzo, "Quantum-assisted indoor localization for uplink mm-Wave and downlink visible light communication systems," IEEE Access, vol. 5, pp. 23 327-23 351, Jul. 2017. source

In the paper

“To address this, a novel approach of quantum routing mechanisms by executing quantum circuits in parallel at the intermediate nodes has been recently proposed in [134] and it has been shown that the proposed quantum routing approach is independent of the number of routing hops and is closer to the optimum in terms of time taken to teleport a quantum state.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [134] 134. J. Du, "Terahertz essential devices technique research based on quantum communication," in proc. of IEEE Advanced Information Technology, Electronic and Automation Control Conf., Oct. 2018, pp. 1927-1930.

In the paper

“Also, another promising application area is the quantum Internet, which enables the transmission of Qubits from one quantum computer to another [119].”

What the sources say

• ?Not checked — no full text of the sourcecites [119] 119. S. Imre, "Quantum communications: Explained for communication engi- neers," IEEE Commun. Mag., vol. 51, no. 8, pp. 28-35, Aug. 2013. source

In the paper

“Another important application of quantum communications is quantum teleportation, which utilizes the quantum entanglement principle to transfer a particular quantum state to another place with the quantum devices by using the classical bits rather than the quantum bits [128].”

What the sources say

• ?Not checked — no full text of the sourcecites [128] 128. K. Inoue, "Quantum key distribution technologies," IEEE J. Sel. Topics Quantum Electron., vol. 12, no. 4, pp. 888-896, Jul. 2006. source

In the paper

“The implementation of Quantum SVMs (QSVMs) has been presented in [185], where an exponential speed-up of QSVMs over classical SVMs has been reported.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [185] 185. A. Kapoor, N. Wiebe, and K. Svore, "Quantum perceptron models," in proc. of Advances in Neural Information Processing Systems, 2016, pp. 3999-4007.

In the paper

“However, this provision is only limited to perform different isolated tasks of optimization, control, and management nature.”

In the paper

“Superconducting qubits and ion traps are the popular forms of the current architecture for quantum computation, which are arrays of interacting qubits that are continuously controlled via external pulses to implement the desired operations [192].”

What the sources say

• ?Not checked — no full text of the sourcecites [192] 192. M. K. Saggi and S. Jain, "A survey towards an integration of big data analytics to big insights for value-creation," Information Processing & Management, vol. 54, no. 5, pp. 758-790, 2018. source

In the paper

“Long-distance quantum communication can be suitably realised through the physical platform of photons, where an open research problem is the loss of photons in quantum channels.”

In the paper

“In the concept of fully integrated smart cities in 6G and beyond communications, the voluminous amount of data instantaneously produced from the massive amount of IoT devices (almost everything connected to the network) can only be effectively utilized to provide instantaneous (runtime) services if a very high computational capability is leveraged to the system.”

In the paper

“Intelligently caching the data at serving stations (e.g., BSs) in the proximity of mobile users can enable offloading heavy traffic from the network backhaul and reducing the network latency for popular contents, through an instantaneous service from the network edge.”

In the paper

“Moreover, the framework of DL has a recognized potential in multi-objective optimization, see e.g., [203].”

What the sources say

• ?Not checked — no full text of the sourcecites [203] 203. H. Yetgin, K. T. K. Cheung, and L. Hanzo, "Multi-objective routing optimization using evolutionary algorithms," in IEEE Wireless Commun. and Networking Conf. IEEE, 2012, pp. 3030-3034. source

In the paper

“Via numerical results, it was shown that quantumassisted precoding provides better BER performance as compared to the conventional PSO algorithm, while keeping the same computational complexity.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [38] 38. P. Botsinis, D. Alanis, Z. Babar, H. V. Nguyen, D. Chandra, S. X. Ng, and L. Hanzo, "Quantum-aided multi-user transmission in non-orthogonal multiple access systems," IEEE Access, vol. 4, pp. 7402-7424, Jul. 2016.

In the paper

“For example, in the paradigm of integrating ML in almost everything has recognized exposure of new types of privacy and security vulnerabilities, while the current understanding of these aspects is very limited [208].”

What the sources say

• ?Not checked — no full text of the sourcecites [208] 208. S. Verma, Y. Kawamoto, Z. M. Fadlullah, H. Nishiyama, and N. Kato, "A survey on network methodologies for real-time analytics of massive IoT data and open research issues," IEEE Commun. Surveys Tuts., vol. 19, no. 3, pp. 1457-1477, Apr. 2017. source

In the paper

“This is because the simulation of a quantum computer on a classical computer is a computationally hard problem.”

In the paper

“Also, regarding the transmission of quantum data over a quantum channel, the achievable rate is given by the Lloyd-Shor-Devetak (LSD) coding theorem [124,137].”

What the sources say

• ?Not checked — no full text of the sourcecites [124] 124. C. H. Bennett, P. W. Shor, J. A. Smolin, and A. V. Thapliyal, "Entanglement-assisted capacity of a quantum channel and the reverse shannon theorem," IEEE Trans. Inf. Theory, vol. 48, no. 10, pp. 2637- 2655, Oct. 2002. source
• ?Not checked — no full text of the sourcecites [137] 137. A. S. Holevo, "The capacity of the quantum channel with general signal states," IEEE Trans. Inf. Theory, vol. 44, no. 1, pp. 269-273, Jan. 1998. source

In the paper

“The role of quantum cryptography, through its features of generating secret keys to multiple legitimate parties in a Vernam one-time pad cryptosystems, is highly recognized as a strong potential for the future of security/privacy [127].”

What the sources say

• ?Not checked — no full text of the sourcecites [127] 127. V. Sharma and S. Banerjee, "Analysis of quantum key distribution based satellite communication," in proc. of Int. Conf. on Computing, Commun. and Networking Technol., Jul. 2018, pp. 1-5. source

In the paper

“For various learning tasks, QML algorithms can provide exponential speed-ups over classical ML algorithms [42], involving large dimensional data, in both supervised and unsupervised learning approaches.”

What the sources say

• ?Not checked — no full text of the sourcecites [42] 42. S. Lloyd, M. Mohseni, and P. Rebentrost, "Quantum algorithms for supervised and unsupervised machine learning," arXiv preprint arXiv:1307.0411, 2013. source

In the paper

“However, this approach will face scalability issues even if they are highly capable of maintaining quantum coherence for a longer duration.”

In the paper

“Moreover, to implement such quantum repeaters, the decoherence effects imposed by the quantum channels need to be dealt with.”

In the paper

“This issue can be addressed by the QKD, also called the quantum cryptography, which provides a secret key to two legitimate parties in a Vernam one-time pad cryptosystem [127].”

What the sources say

• ?Not checked — no full text of the sourcecites [127] 127. V. Sharma and S. Banerjee, "Analysis of quantum key distribution based satellite communication," in proc. of Int. Conf. on Computing, Commun. and Networking Technol., Jul. 2018, pp. 1-5. source

In the paper

“The development of highly consistent and controllable qubits and quantum logical operations is a fundamental need in the realization of reliable large-scale quantum computers, where quantum error correction methods can be used in bringing improvements in system reliability.”

In the paper

“More specifically, fault-tolerant quantum computers are required to solve harder problems.”

In the paper

“In terms of practical implementation, a quantum annealing chipset is already commercially available from the company D-Wave1 [135].”

What the sources say

• ?Not checked — no full text of the sourcecites [135] 135. S.-T. Cheng, C.-Y. Wang, and M.-H. Tao, "Quantum communication for wireless wide-area networks," IEEE J. Sel. Areas Commun., vol. 23, no. 7, pp. 1424-1432, Jul. 2005. source

In the paper

“Parallelization can partially facilitate the resolution by allowing the simulation of more qubits in less time.”

In the paper

“Alos, the scope of DL and evolutionary game theory for dynamic spectrum access in a cognitive radio network has been discussed in [237].”

What the sources say

• ?Not checked — no full text of the sourcecites [237] 237. S. Singhal, S. Gupta, and A. Sinha, "Role of artificial intelligence in cognitive radio networks," in proc. of Sensing Techniques for Next Generation Cognitive Radio Networks. IGI Global, 2019, pp. 258-279. source

In the paper

“Therefore, ML methods have thoroughly been reviewed for cognitive radios, see e.g., [235].”

What the sources say

• ?Not checked — no full text of the sourcecites [235] 235. Z. Ding, D. W. K. Ng, R. Schober, and H. V. Poor, "Delay minimization for NOMA-MEC offloading," IEEE Signal Process. Lett., vol. 25, no. 12, pp. 1875-1879, Dec. 2018. source

In the paper

“The TDD scheme is usually preferred over FDD in such large-scale multi-antenna systems for the reasons to avoid the tedious task of estimation of the massive amount of downlink channels at the userend, however, this causes the pilot contamination problem imposing limitations over the capacity, which can get severe in high mobility scenarios.”

In the paper

“This popularity of auto-encoding of a single end-to-end communication link in a DNN, is a strong motivation to extend this idea of auto-encoding an entire heterogeneous wireless access network in a QC-assisted very deep ANN to enable an instantaneous unique encoding response for service provisions through hybrid 1∼6G network infrastructures.”

In the paper

“The NOMA principle is also believed to be a convenient method in realizing massive connectivity in the context of IoT in emerging 5G networks [227][228][229].”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [227][228][229]

In the paper

“The user-centric transmission not only overcomes the inter-cell interference encountered in the conventional cellular networks but also provides macro-scale diversity [219].”

What the sources say

• ?Not checked — no full text of the sourcecites [219] 219. X.-Z. Cui, X.-L. Yin, H. Chang, Y.-L. Guo, Z.-J. Zheng, Z.-W. Sun, G.-Y. Liu, and Y.-J. Wang, "Analysis of an adaptive orbital angular momentum shift keying decoder based on machine learning under oceanic turbulence channels," Optics Communications, vol. 429, pp. 138-143, 2018. source

In the paper

“The aforementioned indications assert that the truly intelligent self-sustainability in communication networks will be materialized in 6G and beyond communication with QCassisted ML and QML as the potential enablers.”

In the paper

“Moreover, the advanced ML concepts of DL, online-learning, transfer learning, and auto-encoder can together help to overcome these limitations.”

In the paper

“Both NOMA and MEC being intelligence and computing power demanding technologies, QML can be seen as a strong potential enabler.”

In the paper

“The hardware support for such SIC at user nodes is also released as NOMA-chipset [226].”

What the sources say

• ?Not checked — no full text of the sourcecites [226] 226. Y. Ding and B. D. Rao, "Dictionary learning-based sparse channel representation and estimation for FDD massive MIMO systems," IEEE Trans. Wireless Commun., vol. 17, no. 8, pp. 5437-5451, Aug. 2018. source

In the paper

“The antenna tweaking parameter for configuring antennas at the userside opens the possibility of intelligently manipulating them through ML for overall performance optimization.”

In the paper

“The role of AI in enabling cognitive radio networks has been strongly endorsed in [236].”

What the sources say

• ?Not checked — no full text of the sourcecites [236] 236. C. Clancy, J. Hecker, E. Stuntebeck, and T. O'Shea, "Applications of machine learning to cognitive radio networks," IEEE Wireless Commu- nications, vol. 14, no. 4, pp. 47-52, Aug. 2007. source

In the paper

“All the tasks for operating cognitive radios, e.g., sensing spectrum gaps, network (spatial) interference analysis, CSI estimation, power control, and dynamic resources management, etc, have natural connections with the deals offered by ML methods.”

In the paper

“The training of DNNs is usually performed through backpropagating the gradient of the loss function, however, the unavailability of prior knowledge of channel statistics, in end-to-end DNN systems, prevents the back-propagation of gradients.”

In the paper

“The 5G era is expected to lay the foundation of intelligent communication networks [54] by introducing some basic AI based operations, such as intelligent resource management, intelligent management of services provision, intelligent control, etc.”

What the sources say

• ?Not checked — no full text of the sourcecites [54] 54. R. Li, Z. Zhao, X. Zhou, G. Ding, Y. Chen, Z. Wang, and H. Zhang, "Intelligent 5G: when cellular networks meet artificial intelligence," IEEE Wireless Communications, vol. 24, no. 5, pp. 175-183, Oct. 2017. source

In the paper

“With the growing size and complexity of communication networks, the future of cognitive radios seems to be in the sensing and optimization through advanced ML methods assisted with quantum speed-ups.”

In the paper

“The software-defined (fluid) antennas are expected to provide rich diversity at the user nodes, of similar level which M-MIMO systems can provide at the BS side.”

In the paper

“These studies involve the assumption of prior available channel statistics (i.e., available differentiable channel model).”

In the paper

“The processing and learning capability required for manipulating multi-dimensional, large-sized, and highly-dynamic (V2V) streaming data in such large-scale software-defined multi-antenna systems; is far beyond the reach of classical computational and learning methods.”

In the paper

“The amount of available configurable parameters at a communication node is rapidly increasing, which is projected to be 2000+ parameters in a typical 5G node [54].”

What the sources say

• ?Not checked — no full text of the sourcecites [54] 54. R. Li, Z. Zhao, X. Zhou, G. Ding, Y. Chen, Z. Wang, and H. Zhang, "Intelligent 5G: when cellular networks meet artificial intelligence," IEEE Wireless Communications, vol. 24, no. 5, pp. 175-183, Oct. 2017. source