Achieving Geometric Convergence for Distributed Optimization Over Time-Varying Graphs
Open accessThe DOI matches its Crossref record; 89 of the 89 references it cites were found in a registry or an index; 10 of the 34 sources its claims cite were checked for support.
Abstract
This paper considers the problem of distributed optimization over time-varying graphs. For the case of undirected graphs, we introduce a distributed algorithm, referred to as DIGing, based on a combination of a distributed inexact gradient method and a gradient tracking technique. The DIGing algorithm uses doubly stochastic mixing matrices and employs fixed step-sizes and, yet, drives all the agents' iterates to a global and consensual minimizer. When the graphs are directed, in which case the implementation of doubly stochastic mixing matrices is unrealistic, we construct an algorithm that incorporates the push-sum protocol into the DIGing structure, thus obtaining the Push-DIGing algorithm. Push-DIGing uses column stochastic matrices and fixed step-sizes, but it still converges to a global and consensual minimizer. Under the strong convexity assumption, we prove that the algorithms converge at R-linear (geometric) rates as long as the step-sizes do not exceed some upper bounds. We establish explicit estimates for the convergence rates. When the graph is undirected it shows that DIGing scales polynomially in the number of agents. We also provide some numerical experiments to demonstrate the efficacy of the proposed algorithms and to validate our theoretical findings.
Overview
CoChat Summary
TL;DR
The paper introduces DIGing for distributed optimization over time-varying undirected graphs and Push-DIGing for time-varying directed graphs, both using fixed step-sizes and gradient tracking. Under strong convexity and smoothness, it proves global R-linear (geometric) convergence and gives explicit rate bounds.
Key findings
- DIGing uses doubly stochastic mixing matrices and fixed step-sizes; under Assumptions 1, 4, and 5, it converges to the unique optimum at a global R-linear rate \(O(\lambda^k)\) for step-sizes \(\alpha \in (0, 1.5(1-\delta)^2 \bar{\mu} J_1]\).
- Push-DIGing incorporates the push-sum protocol with column stochastic matrices and fixed step-sizes; under Assumptions 4–7, it also converges at a global R-linear rate \(O(\lambda^k)\) for \(\alpha \in (0, 1.5(1-\delta)^2 \bar{\mu} J_2]\).
- For undirected graphs, Corollary 11 gives iteration complexity \(O(\tau^{-2} B^3 n^{4.5} \bar{\kappa}^{1.5} \ln(1/\varepsilon))\), polynomial in the number of agents \(n\). For lazy Metropolis mixing on time-invariant connected undirected graphs, Corollary 12 gives rate \(\lambda = 1 - 1/(161312\, n^{4.5} \bar{\kappa}^{1.5})\) and iteration complexity \(O(n^{4.5} \bar{\kappa}^{1.5} \ln(1/\varepsilon))\).
- Numerical experiments with \(n=12\), \(m_i=1\), \(p=3\), and Huber loss show DIGing and variants have R-linear rates, while the push-gradient method has only sublinear rate; DIGing-ATC is faster in iterations than DIGing but has higher per-iteration communication cost, and Push-DIGing can be as fast as DIGing-ATC at half communication cost per iteration.
Methods
- The paper studies distributed convex optimization \(\min \sum_{i=1}^n f_i(x)\) over time-varying undirected and directed graphs, assuming smoothness with Lipschitz gradients and strong convexity of at least one local function.
- It proposes DIGing: \(x(k+1)=W(k)x(k)-\alpha y(k)\), \(y(k+1)=W(k)y(k)+\nabla f(x(k+1))-\nabla f(x(k))\), and Push-DIGing, which adds push-sum variables \(u,v,V\) and uses column stochastic matrices.
- Convergence analysis uses the small-gain theorem with an ergodic norm and a cycle of arrows \(q \to z \to \check{y} \to \check{x} \to q\); numerical experiments compare push-gradient, DIGing, DIGing-ATC, and Push-DIGing on time-invariant directed, time-varying undirected, and time-varying directed graphs.
Limitations
- The paper states an open question is obtaining improved estimates on convergence rates, especially scaling with the number of agents \(n\), and extensions to local constraints and couplings among agents.
- Step-size selection requires agents to agree on common upper bounds on \(n\) and \(B\), plus lower \(\bar{\mu}\) and upper \(L\); the paper assumes all agents use the same step-size, though it notes uncoordinated step-sizes are possible in subsequent work.
- The paper states its bounds on step-sizes are often conservative and uses hand-optimized step-sizes in experiments; the numerical experiments use a small network of \(n=12\), so the empirical validation is limited in scale.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
9 Valid · 0 Not valid · 10 Not checked · 37 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This compares favorably to some of the existing literature, e.g., [70,76], which require detailed spectral information about the network for step-size selection.”
What the sources say
- Supportedcites [70] C. Xi and U. Khan, On the Linear Convergence of Distributed Optimization over Directed Graphs , preprint,arXiv:1510.02149, 2015. source
The part of the sentence this citation is offered for — that the cited work's step-size selection needs detailed spectral information about the network — is borne out by our copy: Theorem 1 (p. 13) sets α_min and α_max in terms of λ_min(M+M^⊤), d_- and d_∞, and the discussion after it (p. 14) states outright that 'the values of α and ᾱ need the knowledge of the network topology, which may not be available in a distributed manner,' flagging avoidance of that global knowledge as an open question. Section VI (p. 22) demonstrates the requirement in practice, computing λ_max/λ_min of H_i^⊤H_i, the degree constants and λ_min(M+M^⊤) to estimate the interval [9.6×10^-4, 0.26] before picking α = 0.1. The comparative framing ('compares favorably') is the citing authors' own conclusion and is not attributed to this source, so it is not judged here. (LLM, conf=0.85)
- Partly supportedcites [76] J. Zeng and W. Yin, ExtraPush for Convex Smooth Decentralized Optimization over Directed Networks , preprint,arXiv:1511.02942, 2015. sourceExperimental — accuracy not yet measured
Theorem 4 (p. 8) gives a condition on the step size, and the discussion on p. 9 relates the convergence bound to eigenvalues of network-derived matrices. But p. 9 calls the step-size condition sufficient, not necessary, and §6.1 (p. 12) selects step sizes by hand. The paper therefore supports a spectral dependence in its theoretical bound, not the categorical claim that step-size selection requires detailed spectral information. (LLM, conf=0.79)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Proposition 1 ( [62]). Assume null { I -W } = span { 1 } where W ∈ R n × n . If x ∗ satisfies conditions: (i) x ∗ = Wx ∗ (consensus), (ii) 1 ⊤ ∇ f ( x ∗ ) = 0 (optimality), then the rows of x ∗ are the same as each other and the transpose of each row is an optimal solution of the problem (1) .”
What the sources say
- Supportedcites [62] matched to “EXTRA: An Exact First-Order Algorithm for Decentralized Consensus Optimization” (2015) source
The paper’s discussion of Proposition 2.1 (p. 6) says that, under null{I−W} = span{1}, a point satisfying its consensus and optimality conditions has identical rows that solve problem (1.1). Lemma 3.1’s proof (p. 9) identifies 1ᵀ∇f(x*) = 0 as the optimality condition. The citing proposition paraphrases those conditions and conclusion. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“While the subgradient-push eliminates the requirement of graph balancing [22], it suffers from a slow sublinear 2 convergence rate even for strongly convex smooth functions due to its employment of diminishing step-size [41].”
What the sources say
- Supportedcites [41] A. Nedić and A. Olshevsky, Stochastic Gradient-Push for Strongly Convex Functions on Time-Varying Directed Graphs , preprint, arXiv:1406.2075, 2014. source
The paper reports an O((ln t)/t) rate for strongly convex functions with Lipschitz gradients using an appropriately chosen diminishing step-size (main-result passage; page unavailable). This is sublinear, although the Introduction and Conclusion emphasize that it improves on the earlier O((ln t)/√t) rate and is close to the best achievable rate in the noisy setting; “slow” is the citing authors’ characterization. (LLM, conf=0.85)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Assumption 2 is typical for many results in multi-agent coordination and distributed optimization [42]. It is much weaker than the assumption of every G un ( k ) being connected.”
What the sources say
- Supportedcites [42] matched to “Distributed Optimization Over Time-Varying Directed Graphs” (2015) source
In II. Problem, Algorithm and Main Results (p. 2), the paper calls its connectivity assumption typical and says it is considerably weaker than requiring each graph to be connected, because the necessary edges can appear over time. (LLM, conf=0.91)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When the graph is undirected it shows that DIGing scales polynomially in the number of agents.”
- Also stated in: § ACHIEVING GEOMETRIC CONVERGENCE FOR DISTRIBUTED OPTIMIZATION OVER TIME-VARYING GRAPHS ∗ · p. 4 · plain
“Moreover, we show that the convergence rate of DIGing scales polynomially in the total number of agents in the network.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on 2 assumptions not shown here.
- Also stated in: § ACHIEVING GEOMETRIC CONVERGENCE FOR DISTRIBUTED OPTIMIZATION OVER TIME-VARYING GRAPHS ∗ · p. 4 · plain
25 more claims
Create a free account to see all 56 claims and to verify the 2 references still open.
- It may state two separate claims as one. 13
- Probably not a substantial claim. 14
- The quote only partly states this claim. 1
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Algorithm 2: Push-DIGing · p. 21
“N out i ( k ) is the set of agents that can receive the information from agent i (in-neighbors of agent i ) at time k .”
§ Algorithm 2: Push-DIGing · p. 21
“N out i ( k ) = { j ∣ ∣ ( - → i, j ) ∈ A ( k ) }”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Algorithm 1: DIGing · p. 10
“the gradients are close to zero in the neighborhood of the optimal point”
§ ACHIEVING GEOMETRIC CONVERGENCE FOR DISTRIBUTED OPTIMIZATION OVER TIME-VARYING GRAPHS ∗ · p. 5
“not necessarily a point where ∇ f i ( x ) = 0 ∀ i”
Citations
Cites
Boyd, S., Parikh, N., Chu, E., Peleato, B., & Eckstein, J. (2011) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000016 · Cited by 17,075
Boyd, S. (2010) · now publishers, Inc. eBooks · doi:10.1561/9781601984616 · Cited by 12,677
Distributed Subgradient Methods for Multi-Agent Optimization
Nedic, A., & Ozdaglar, A. (2009) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2008.2009515 · Cited by 3,719 · cited 2×
“References [60–62] use a difference structure to cancel the steady state error in decentralized gradient descent [47, 75], thereby developing the algorithm EXTRA and its proximal-gradient variant.”
MethodFast linear iterations for distributed averaging
Xiao, L., & Boyd, S. (2004) · Systems & Control Letters · doi:10.1016/j.sysconle.2004.02.022 · Cited by 2,706
Markov Chains and Mixing Times
Levin, D., Peres, Y., & Wilmer, E. (2008) · doi:10.1090/mbk/058 · Cited by 2,241
Distributed asynchronous deterministic and stochastic gradient optimization algorithms
Tsitsiklis, J., Bertsekas, D., & Athans, M. (1986) · IEEE Transactions on Automatic Control · doi:10.1109/tac.1986.1104412 · Cited by 1,967
Gossip-based computation of aggregate information
Kempe, D., Dobra, A., & Gehrke, J. (2004) · 44th Annual IEEE Symposium on Foundations of Computer Science, 2003. Proceedings. · doi:10.1109/sfcs.2003.1238221 · Cited by 1,480 · cited 4×
“Instead, if every agent knows its out-degree, it is possible for the agents to construct a column stochastic matrix C and perform the following recursions with initialization u(0) = x(0) and v(0) = 1 to achieve the average (push-sum protocol [27]):”
BackgroundMethodInfluentialEXTRA: An Exact First-Order Algorithm for Decentralized Consensus Optimization
Shi, W., Ling, Q., Wu, G., & Yin, W. (2015) · SIAM Journal on Optimization · doi:10.1137/14096668x · Cited by 1,261
Self-citationDistributed Optimization Over Time-Varying Directed Graphs
Nedic, A., & Olshevsky, A. (2015) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2014.2364096 · Cited by 1,187
Self-citationDistributed average consensus with least-mean-square deviation
Xiao, L., Boyd, S., & Kim, S.-J. (2007) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2006.08.010 · Cited by 1,107 · cited 1×
“Some examples include distributed averaging [7,50,72], distributed spectrum sensing [1], formation control [49, 58], power system control [21, 55], statistical inference and learning [20, 45, 53].”
Background
Research works printed but not linked: 9
- Chen A., 2012, Allerton, 601
- Di Lorenzo P., 2015, Proceedings of the IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, 229
- Forero P., 2010, J. Mach. Learn. Res., 59, 1663
- Mokhtari A., 2016, J. Mach. Learn. Res., 17, 1
- Nedić A., 2008, UK, 3
- A. Olshevsky, Efficient Information Aggregation Strategies for Distributed Control and Signal Processing , Ph.D. thesis, Massachusetts Institute of Technology, 2010.
- Rabbat M., 2004, New York, 20
- Ram S., 2009, Proceedings of INFOCOM, 3001
- Sayed A., 2013, New York, 323
Cited by
Items 601–700 of 1,070
Zhang, K., Yang, Z., Liu, H., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.02783 · Cited by 4
“Since we consider a more general setting with a time-varying communication network, several recent work [53, 54, 55, 56] may apply.”
BackgroundMethodResultInfluentialA Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!Preprint
Kovalev, D. A., Koloskova, A., Jaggi, M., Richtárik, P., & Stich, S. U. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.01697 · Cited by 75
“…not attaining linear convergence rates, have been pointed out for instance in (Shi et al., 2015) and techniques such as EXTRA (Shi et al., 2015) and gradient tracking (Nedić et al., 2017) have been developed to achieve linear convergence on strongly convex problems with primal methods as well.”
BackgroundMethodA fast randomized incremental gradient method for decentralized non-convex optimizationPreprint
Xin, R., Khan, U. A., & Kar, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.03853 · Cited by 4
Decentralized Composite Optimization with CompressionPreprint
Li, Y., Liu, X., Tang, J., Yan, M., & Yuan, K. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.04448 · Cited by 13
“…EXTRA (Shi et al., 2015a), NIDS (Li et al., 2019), D2 (Tang et al., 2018b), Exact Diffusion (Yuan et al., 2018b), NEXT (Di Lorenzo and Scutari, 2016), DIGing (Nedic et al., 2017), Harnessing (Qu and Li, 2017), SONATA (Scutari and Sun, 2019), GSGT (Pu and Nedić, 2020), OPTRA(Xu et al., 2020a), etc.”
Distributed Proximal Algorithms for Nonsmooth Optimization: Unified Convergence Analysis
Huang, Y., Cui, S., Sun, J. Q., & Meng, Z. (2025) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2025.3576271 · Cited by 5
“To solve these two classes of distributed optimization problems, many discrete-time and continuous-time algorithms have been proposed in [2], [3], [4], [5], [6], [7], [8], and [9].”
Distributed online aggregation optimization with time‐varying constraints
Zhang, X., Mo, L., & Yang, T. (2025) · Asian Journal of Control · doi:10.1002/asjc.3747 · Cited by 3
Distributed Optimization Under Information Constraints: A Survey
Liu, S., Hua, Y., Han, Q., & Xie, L. (2026) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2025.3645932 · Cited by 7
“Toaccelerateconvergence,laterstudiesintroduced gradienttracking(GT)methods,suchasEXTRA[6]andDIG-ing[7],alongwiththeirrefinedvariantslikeNIDS[8]and exactdiffusion[9],[10].”
Linearly Convergent Second-Order Distributed Optimization Algorithms
Qu, Z., Li, X., Li, L., & Hong, Y. (2024) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2024.3360287 · Cited by 4
“In[36] and[37],theATCstrategyisappliedtoimprovetheperformanceof trackingstrategy[6].”
Multi-consensus decentralized primal-dual fixed point algorithm for distributed learning
Tang, K., Liu, W., & Mao, X. (2024) · Machine Learning · doi:10.1007/s10994-024-06537-8 · Cited by 4
“We refer the reader to Olshevsky (2017) for a variant of (8).”
Distributed Conjugate Gradient Method via Conjugate Direction Tracking
Shorinwa, O., & Schwager, M. (2024) · Proceedings of the ... American Control Conference/Proceedings of the American Control Conference · doi:10.23919/acc60939.2024.10644588 · Cited by 9
“…section, we examine the performance of our distributed conjugate gradient method (DC-Grad) in comparison to other existing distributed optimization algorithms, namely: DIGing-ATC [14], C-ADMM [17], AB /Push-Pull [16], and ABm [13], which utilizes momentum acceleration to achieve faster convergence.”
BackgroundMethodInfluentialOn the Optimal Communication Weights in Distributed Optimization Algorithms
Colla, S., & Hendrickx, J. M. (2024) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2024.10.107 · Cited by 3
Distributed Stochastic Learning for Composite Sharing Optimization in Consumer Electronics
Lü, Q., Dai, X., Zhang, W., & Ezilarasan, M. R. (2024) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2024.3506915 · Cited by 3
“Related known approaches can be found in the relevant works [15], [16], [17], [18], [19], [20], [21], [22], [23], [24].”
InfluentialCui, G., Jia, Q., & Guan, X. (2025) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2025.3534639 · Cited by 4
“Condition (3.3) holds if and only if the communication graph is connected [24].”
InfluentialMomentum-Based Distributed Economic Dispatch of Smart Grids in Time-Varying Directed Networks
Zhang, K., Lü, Q., Zhou, Q., et al. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3561195 · Cited by 3
“In addition, z ki with the initialization ensures that the al-gorithm possesses the capability to converge to the exact optimal solution under constant step sizes (see [54] for more details).”
Ni, X., Wei, Y., Tao, M., et al. (2025) · IEEE Transactions on Industrial Cyber-Physical Systems · doi:10.1109/ticps.2025.3615894 · Cited by 3
Aggregating Stochastic Gradients in Distributed Optimization
Doan, T. T. (2018) · Proceedings of the ... American Control Conference/Proceedings of the American Control Conference · doi:10.23919/acc.2018.8431394 · Cited by 5
“In particular, there are three widely-studied algorithms for distributed optimization, namely, the alternating direction method of multipliers (ADMM) [1], [9], distributed dual averaging [5], [19], and distributed gradient algorithms [11], [13], [14], [16]; to name just a subset of the existing literature.”
MethodAN EXPONENTIALLY CONVERGENT ALGORITHM FOR LEARNING UNDER DISTRIBUTED FEATURES
Ying, B., Yuan, K., & Sayed, A. H. (2018) · IEEE Data Science Workshop (DSW) · doi:10.1109/dsw.2018.8439886 · Cited by 2
“The dynamic average consensus method has been adopted in the distributed optimization algorithms to track the average of gradients, see [15–17].”
MethodConsensus over evolutionary graphs
Smytnakis, M., Freris, N. M., & Tembiné, H. (2018) · European Control Conference (ECC) · doi:10.23919/ecc.2018.8550392 · Cited by 3
“In this paper, we will study consensus over time-varying graphs as abstracted by a time-varying Laplacian Lt, dictated by agents’ randomized decisions; that is to say, Lt = L(xt, Et−) is a random matrix that depends on both the state at time t and the topology ‘right before’ time t; consequently, the analysis in exist…”
MethodAccelerated Distributed Primal-Dual Dynamics Using Adaptive Synchronization
Bansode, P., Kosaraju, K. C., Wagh, S. R., Pasumarthy, R., & Singh, N. M. (2019) · IEEE Access · doi:10.1109/access.2019.2937448 · Cited by 2
“Lately, the algorithms such as distributed gradient (sub-gradient) methods have been widely re-studied with the objective of improvement in the rate of convergence, see [10, 14, 23–25].”
MethodImproving the Convergence of Distributed Gradient Descent via Inexact Average Consensus
Du, B., Zhou, J., & Sun, D. (2020) · Journal of Optimization Theory and Applications · doi:10.1007/s10957-020-01648-3 · Cited by 4
Distributed Optimization With Uncertain Communications
Rezaeinia, P., Gharesifard, B., & Linder, T. (2024) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2024.3495456 · Cited by 2
“This selective participation differentiates an effective communication network graph from conventional network graphs that do not filter connections based on these criteria.”
Shi, Z., & Zhou, C. (2020) · Optimal Control Applications and Methods · doi:10.1002/oca.2643 · Cited by 3
A Heavy-Ball Distributed Optimization Algorithm over Digraphs with Row-Stochastic Matrices
Shen, Y., & Yang, S. (2020) · Chinese Control Conference (CCC) · doi:10.23919/ccc50068.2020.9188385 · Cited by 4
“Subsequently, many other push-sum based algorithms are developed [11–13].”
BackgroundMethodColla, S., & Hendrickx, J. M. (2022) · Proceedings of the IEEE Conference on Decision & Control, including the Symposium on Adaptive Processes · doi:10.1109/cdc51059.2022.9993346 · Cited by 3
“This class of decentralized methods, denotedMD, includes many algorithms such as DGD [2], EXTRA [3], DIGing [4], and NIDS [5], to name a few.”
BackgroundMethodInfluentialOptimizing Decentralized Learning with Local Heterogeneity using Topology Morphing and Clustering
Abebe, W. M., & Jannesari, A. (2023) · IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid) · doi:10.1109/ccgrid57682.2023.00041 · Cited by 5
An accelerated exact distributed first-order algorithm for optimization over directed networks
Wang, Z., Wang, C., Wang, J., Hu, J., & Li, H. (2023) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2023.08.015 · Cited by 3
Zhao, S., & Liu, Y. (2023) · Automatica · doi:10.1016/j.automatica.2023.111352 · Cited by 7
“Statistical inference is a core topic in statistics, and confidence regions have been widely used to quantify the uncertainty in the estimation of model parameters.”
NPGA: A Unified Algorithmic Framework for Decentralized Constraint-Coupled Optimization
Li, J., & Su, H. (2024) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2024.3354882 · Cited by 5
“It is well known that (P2) has been widely studied in recent years and numerous effective algorithms has been proposed to solve it, including DIGing [23], [24]; exact first-order algorithm (EXTRA) [25]; decentralized linearized alternating direction method of multipliers (DLM) [26]; proximal primal–dual diffusion (P2D2…”
Delay-Agnostic Asynchronous Distributed Optimization
Wu, X., Liu, C., Magnússon, S., & Johansson, M. (2023) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc49753.2023.10383647 · Cited by 14
“Distributed algorithms for solving Problem (1) include the distributed subgradient method [19], DGD [17], distributed gradient-tracking-based algorithm [20], distributed dual averaging [21], and PG-EXTRA [22].”
MethodRan, L., Li, H., Hu, J., Lü, Q., Wang, Z., Li, Z., & Chen, G. (2024) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2023.3285865 · Cited by 4
“…linear convergence rates of both algorithms are also obtained with additional assumptions, that is, objective functions satisfy a certain quadratic growth condition (weaker than the strong convexity condition in [22], [23], [25], [26], [27], and [28]) or a piecewise linear-quadratic condition.”
BackgroundMethodInfluentialNetwork-GIANT: Fully Distributed Newton-Type Optimization via Harmonic Hessian Consensus
Maritan, A., Sharma, G., Schenato, L., & Dey, S. (2023) · IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464472 · Cited by 8
“The networked setting has been explored with gradient-based optimization in a number of works, e.g. [8]– [11], whereas networked version of the Alternating Direction Method of Multipliers (ADMM) have been explored in [12], [13].”
MethodLuan, M., Wen, G., & Yang, T. (2024) · International Journal of Robust and Nonlinear Control · doi:10.1002/rnc.7666 · Cited by 3
“Clearly, according to Assumption 1, 𝑀(𝑘) is also doubly stochastic.”
Li, B., & Chi, Y. (2025) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2025.3526081 · Cited by 23
Mu, C., Ding, T., Chang, X., Zhu, S., Xue, Y., Han, Z., & Shahidehpour, M. (2025) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2025.3572363 · Cited by 4
“For globally coupled constraints, both LPGT and gradient tracking-based methods ([33], [34], [35], [36], [37]) require the exchange of neighboring variables with weighted multiplication at each iteration, resulting in a computational complexity of O ( | I | l 1 ) , where l 1 is the number of…”
A Convergence-preserving Data Integrity Attack on Distributed Optimization Using Local Information
Ding, T., Xu, Q., Zhu, S., & Guan, X. (2020) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc42340.2020.9304245 · Cited by 5
“The algorithm (4) is proved to converge to the optimal solution x ∗ linearly [7], [10], where x ∗ = 1 ( x ∗ ) T , x ∗ is the optimal solution of (1) and satisfies”
BackgroundA Distributed Proximal Primal-Dual Algorithm for Nonsmooth Optimization with Coupling Constraints
Wu, X., Wang, H., & Lu, J. (2020) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc42340.2020.9303937 · Cited by 2
D2L: Decentralized dictionary learning over dynamic networks
Daneshmand, A., Sun, Y., Scutari, G., & Facchinei, F. (2017) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2017.7952924 · Cited by 3
“Most of the literature on distributed multi-agent optimization deals with convex, unconstrained optimization problems [8–10] over undirected, static graphs [11–13].”
BackgroundA dual splitting algorithm for distributed resource allocation problems
Xu, J., Zhu, S., Soh, Y. C., et al. (2017) · Proceedings of the IEEE Conference on Decision & Control, including the Symposium on Adaptive Processes · doi:10.1109/cdc.2017.8264349 · Cited by 3
Ren, J., & Haupt, J. (2020) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2020.2996374 · Cited by 5
“One interesting direction for future research would be to consider the generalization of the proposed method to other decentralized structures like the ones used in [28]–[30].”
MethodTang, Y., Zhu, H., & Lv, X. (2020) · IET Control Theory and Applications · doi:10.1049/cth2.12075 · Cited by 4
“), subject to Ly = 0 with y ,col(y1, ..., yN). There have been some distributed rules to achieve optimal consensus or compute the optimal solution to this problem under directed graphs, e.g. [10] and [24]. However, most of these algorithms require some initialization process. Since there might be disturbances or round…”
BackgroundCheng, H., Li, H., & Wang, Z. (2020) · International Journal of Systems Science · doi:10.1080/00207721.2020.1815098 · Cited by 3
Lan, G., Ouyang, Y., & Zhou, Y. (2023) · SIAM Journal on Optimization · doi:10.1137/20m138956x · Cited by 11
Constrained Distributed Nonconvex Optimization over Time-varying Directed Graphs
He, Z., He, J., Chen, C., & Guan, X. (2020) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc42340.2020.9304164 · Cited by 4
“Primal methods generally combine (sub)gradient descent with consensus, so as to drive local estimates to converge consensually to the globally optimal point in the primal domain [5]–[7].”
BackgroundA General Framework to Distribute Iterative Algorithms With Localized Information Over Networks
Timoudas, T. O., Zhang, S., Magnússon, S., & Fischione, C. (2023) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2023.3279901 · Cited by 3
“…when g ≡ 0 , i.e., smooth convex optimization, the proximal operator is the identity operator and the distributed algorithm reduces to the algorithms proposed in [33], [34], and [35] (the order of the consensus step W varies, but from a dynamical point of view they are essentially equivalent).”
BackgroundMethodInfluentialDistributed Optimization on Directed Graphs Based on Inexact ADMM with Partial Participation
Yi, D., & Freris, N. M. (2023) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc49753.2023.10384030 · Cited by 3
“Push-DIGing [15] is the most popular gradient-based method for directed graphs that achieves linear convergence through gradient tracking (second line in (11)).”
BackgroundMethodNested Distributed Gradient Methods with Stochastic Computation Errors
Iakovidou, C., & Wei, E. (2019) · Allerton Conference on Communication, Control, and Computing · doi:10.1109/allerton.2019.8919853 · Cited by 2
Feng, Y., Zhang, W., Xiong, J., & Li, H. (2020) · IEEE Access · doi:10.1109/access.2020.2978087 · Cited by 3
“Nedic et al. [17] proposed a distributed inexact gradient method to deal with distributed optimization problem with constant step-size (by adopting the gradient tracking technique) under undirected/directed time-varying graphs.”
BackgroundMethodResultInfluentialSecure and Privacy-Preserving ConsensusPreprint
Ruan, M., Gao, H. B., & Wang, Y. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.04491 · Cited by 3
“To analyze the convergence speed of our approach, following the convergence definition in [31] and [32], we define the convergence rate as at least γ ∈ (0, 1) if there exists a positive constant C such that ∥∥x[k]−α1∥∥ ≤ Cγ holds for all k, where α is the average value of the initial states of all nodes in (15).”
MethodLinear convergence in optimization over directed graphs with row-stochastic matricesPreprint
Xi, C., Mai, V. S., Xin, R., et al. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.06160 · Cited by 3
“, [19], [20], [32], [33], [36].”
BackgroundMethodInfluentialDynamics Based Privacy Protection for Average Consensus on Directed GraphsPreprint
Gao, H. B., & Wang, Y. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.02255 · Cited by 5
“Following the the convergence definition in [35] and [36], we define the convergence rate to be at least γ ∈ (0 , 1) if there exists a positive constant C such that π ( k ) α 1 ≤ Cγ k holds for all k , where π ( k ) = [ π 1 ( k ) , . . . (cid:13)(cid:13) (cid:13)(cid:13) Note that a smaller γ means a…”
Zholbaryssov, M., Hadjicostis, C. N., & Domínguez-García, A. D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.07600 · Cited by 3
“:= jN+ i [k]j+1. Let N i:= fj2V: (i;j) 2E(0)g, and d+ i := jN ij+ 1 denotes the nominal out-degree. Regarding the communication model, we also make the following standard assumption (see, e.g., [14], [18]). Assumption 2. When G(0) is undirected (directed), there exists some positive integer Bsuch that the graph S (k+1…”
BackgroundMethodInfluentialDistributed Dual Gradient Tracking for Resource Allocation in Unbalanced NetworksPreprint
Zhang, J., You, K., & Cai, K. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.09937 · Cited by 3
“…) is given by Thus, it follows from (7) that (9) There are several distributed algorithms to solve (8) over unbalanced and directed networks, such as [29, gradient-push], [33, Push-DIGing], [28, D-DGD], [34, DEXTRA], [35], and their asynchronous counterparts [36, AsySPA], [37, APPG] and [38], [39].”
BackgroundLinearly Convergent Algorithm with Variance Reduction for Distributed Stochastic OptimizationPreprint
Lei, J., Yi, P., Chen, J., & Hong, Y. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.03269 · Cited by 2
“Recently, techniques utilizing uncoordinated steplengths have been proposed in [26].”
BackgroundZimmermann, J., Tatarenko, T., Willert, V., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.02854 · Cited by 3
“These methods, published for example in [7], [9] and [12], have The authors are with the Control Methods and Robotics Lab at TU Darmstadt, Germany.”
MethodDC-DistADMM: ADMM Algorithm for Contsrained Distributed Optimization over Directed GraphsPreprint
Khatana, V., & Salapaka, M. V. (2020) · arXiv (Cornell University) · Cited by 5
Distributed ADMM over directed graphs.Preprint
Rokade, K., & Kalaimani, R. K. (2020) · arXiv (Cornell University) · Cited by 3
An Optimal Algorithm for Strongly Convex Minimization under Affine ConstraintsPreprint
Salim, A., Condat, L., Kovalev, D. A., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.11079 · Cited by 3
“Decentralized optimization is a well-known instance of Problem (1), withK a gossip matrix (or its square root), and b = 0 [3,18,21,25,26,27,30,35,37,42].”
BackgroundDistributed Picard Iteration.Preprint
Andrade, F. L., Figueiredo, M. A. T., & Xavier, J. (2021) · arXiv (Cornell University) · Cited by 3
Accelerating Gossip SGD with Periodic Global AveragingPreprint
Chen, Y., Yuan, K., Zhang, Y., Pan, P., Xu, Y., & Yin, W. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.09080 · Cited by 32
“…explicit bias-correction (Shi et al., 2015; Yuan et al., 2019; Li et al., 2019c), gradient tracking (Xu et al., 2015; Di Lorenzo & Scutari, 2016; Nedic et al., 2017; Qu & Li, 2018), coordinate-descent methods (He et al., 2018), and dual acceleration (Scaman et al., 2017; 2018; Uribe et al.,…”
MethodKovalev, D. A., Gasanov, E., Richtárik, P., & Gasnikov, A. V. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.04469 · Cited by 14
“There are a few linearly-convergent algorithms, such as those of Nedic et al. (2017) and Push-Pull Gradient Method of Pu et al. (2020), that assume a primal oracle, and the dual oracle based algorithm PANDA due to Maros and Jaldén (2018).”
BackgroundMethodChen, X. C., Huang, L., He, L., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.04370 · Cited by 3
“, [7], [21], [22]), where graphs GR and GC> are assumed to be strongly connected.”
BackgroundDecentralized Optimization On Time-Varying Directed Graphs Under Communication ConstraintsPreprint
Chen, Y., Hashemi, A., & Vikalo, H. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.13189 · Cited by 5
Essentially Decentralized Conjugate GradientsPreprint
Engelmann, A., & Faulwasser, T. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.12311 · Cited by 7
Distributed ADMM over directed networks.Preprint
Rokade, K., & Kalaimani, R. K. (2021) · arXiv (Cornell University) · Cited by 4
Resilient Distributed Optimization Algorithm With Fixed Step Size Against Malicious Attacks
Cao, L., Gao, W., & Zhao, J. (2025) · IEEE Transactions on Signal and Information Processing over Networks · doi:10.1109/tsipn.2025.3613875 · Cited by 3
“Therefore, we improve the F-local filtering mechanism and propose a novel FTM, then integrate it into the gradient tracking algorithm presented in [25] to construct fixed step size RDO algorithms.”
Golden Ratio Proximal Gradient ADMM for Distributed Composite Convex Optimization
Yin, C., & Yang, J. (2023) · Journal of Optimization Theory and Applications · doi:10.1007/s10957-023-02336-8 · Cited by 5
“These algorithms include EXTRA [37], exact diffusion [45], NIDS [26], and decentralized gradient tracking [13, 18, 30, 41, 42], which have the same computational complexity as the first-order primal algorithms and converge much faster.”
Consensus on Lie Groups for the Riemannian Center of Mass
Kraisler, S., Talebi, S., & Mesbahi, M. (2023) · Proceedings of the IEEE Conference on Decision & Control, including the Symposium on Adaptive Processes · doi:10.1109/cdc49753.2023.10384031 · Cited by 2
“The simplest way to introduce this cost gradient feedback is through Gradient Tracking (GT) [20].”
MethodNew Convergence Analysis of the BEER Algorithm in Decentralized Nonconvex Optimization
Phuong, T. T., & Phong, L. T. (2024) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-031-70011-8_20 · Cited by 2
A Modified Distributed Gradient Dynamics for Multiagent Optimization on Directed Networks
Jahvani, M., & Guay, M. (2025) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2025.3526343 · Cited by 2
“This is in contrast to the algorithms thatworkwithaconstantgain/step-size,e.g.,[13],[14],[18],and [19].”
InfluentialCommunication-Efficient Federated Optimization Over Semi-Decentralized Networks
Wang, H., & Chi, Y. (2025) · IEEE Transactions on Signal and Information Processing over Networks · doi:10.1109/tsipn.2025.3539004 · Cited by 3
“To address this issue, gradient-tracking (GT) techniques [13], [42], [43] have been proposed, which utilizes gossip communication for global gradient estimation leveraging dynamic average consensus [11].”
A Decentralized Primal-Dual Method With Quasi-Newton Tracking
Wang, L., Wu, H., & Zhang, H. (2025) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2025.3547787 · Cited by 2
Distributed Asynchronous Optimization With Inseparable Coupled Constraints and Its Application
Wang, T., Liu, W., Li, Z., Ye, Z., Peng, X., & Du, W. (2025) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2025.3550740 · Cited by 3
Asymmetrically Decentralized Federated Learning
Li, Q., Zhang, M., Yin, N., Yin, Q., Shen, L., & Cao, X. (2025) · IEEE Transactions on Computers · doi:10.1109/tc.2025.3569185 · Cited by 13
“More optimization analysis can be referred to [34], [35], [36], [37], [38].”
A flexible gradient tracking algorithmic framework for decentralized optimization
Berahas, A. S., Bollapragada, R., & Gupta, S. (2025) · Computational Optimization and Applications · doi:10.1007/s10589-025-00685-w · Cited by 3
“Thus, qualitatively, an increase in the number of communication steps, either by increasing n c or p , leads to a reduction in the computation complexity with a natural speed limit of the centralized gradient descent complexity, depicted in the first term of the upper bound (24).”
BackgroundMethodInfluentialConvergence Analysis of EXTRA in Non-Convex Distributed Optimization
Qin, L., & Pu, Y. (2025) · IEEE Control Systems Letters · doi:10.1109/lcsys.2025.3577964 · Cited by 6
“A key breakthrough in distributed optimization was the introduction of the Ex act Firs t -Orde r A lgorithm ( EXTRA ) in [16] and DIGing [17], which both improve on traditional consensus-based methods by achieving exact consensus.”
Liang, L., Huang, X., Xin, R., & Yuan, K. (2025) · SIAM Journal on Optimization · doi:10.1137/24m1657341 · Cited by 14
“Theorem 2 presents the first finite-time non-convex convergence result for Push-DIGing , while previous analyses either focus on the strongly convex case [15, 39] or the asymptotic rate [19].”
InfluentialNPD-SG: A Noise-Resistant Primal-Dual Stochastic Gradient Diffusion Algorithm Over Networks
Wu, J., Zhou, Z., Zhang, S., & Han, H. (2025) · IEEE Transactions on Signal and Information Processing over Networks · doi:10.1109/tsipn.2025.3600760 · Cited by 2
Guo, L., Wang, L., Shi, X., & Cao, J. (2025) · IEEE Transactions on Signal and Information Processing over Networks · doi:10.1109/tsipn.2025.3600766 · Cited by 7
“2110 × ) Prox-DIGing [7], [16] 1966 [4], [16] 959 ↓ (0 .”
BackgroundMethodDistributed stochastic constrained optimization with constant step-sizes via saddle-point dynamics
Huang, Y., Cui, S., Zeng, X., & Meng, Z. (2025) · Automatica · doi:10.1016/j.automatica.2025.112575 · Cited by 2
Li, Y., Wu, C., Chen, Y., Zhang, K., Lü, Q., Deng, S., & Li, H. (2025) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2025.108040 · Cited by 2
Fedele, G., & D’Alfonso, L. (2025) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2025.3620960 · Cited by 2
“Three main approaches exist: Laplacian gradient (LG) methods [27], [28], gradient tracking (GT) strategies [29], and alternating direction method of multipliers (ADMM)-based schemes [30].”
Globally Constrained Decentralized Optimization With Variable Coupling
Wang, D., Wu, X., Ou, Z., & Lu, J. (2025) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2025.3629001 · Cited by 5
“…on consensus optimization, which aims at finding a common decision that minimizes the sum of the local objectives under, mostly, no constraints [4], [5] or rather simple constraints such as set constraints [6], [7], local equality/inequality constraints [8], [9], and equality/inequality…”
InfluentialLi, C., & Wang, Q. (2026) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2026.108403 · Cited by 2
Multi-agent based optimal equilibrium selection with resilience constraints for traffic flow
Liu, P., Korovin, I., Gorbachev, S. V., Gorbacheva, N., & Cao, J. (2022) · Neural Networks · doi:10.1016/j.neunet.2022.08.013 · Cited by 2
A Distributed Convex Optimization Algorithm with Continuous-Time Communication
Jahvani, M., & Guay, M. (2022) · International Symposium on Advanced Control of Industrial Processes · doi:10.1109/adconip55568.2022.9894148 · Cited by 2
“, [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18].”
Scalable Primal Decomposition Schemes for Large-Scale Infrastructure Networks
Engelmann, A., Shin, S., Pacaud, F., & Zavala, V. M. (2025) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2025.3526709 · Cited by 4
“Prominentapproachesareprimalanddualfirst-orderalgorithms, suchasLagrangiandualdecomposition;thealternatingdi-rectionmethodofmultipliers(ADMMs)[3],[4];andprimal (sub)gradient-basedschemes[5],[6].”
Decentralized Gradient-Free Methods for Stochastic Non-smooth Non-convex Optimization
Lin, Z., Xia, J., Deng, Q., & Luo, L. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i16.29697 · Cited by 3
Decentralized Adaptive TD(λ) Learning With Linear Function Approximation: Nonasymptotic Analysis
Zhu, J., Mao, T., Zhang, M., Ge, Q., Wu, Q., & Li, K. (2024) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2024.3382986 · Cited by 2
“Meanwhile, Lin and Ling [23] presented decentralized TD(0) learning methods by leveraging the gradient tracking technique, which is often used to accelerate the rate of convergence in distributed optimization [36], [37], [38].”
MethodDistributed Optimization with Gradient Tracking Over Heterogeneous Delay-Prone Directed Networks
Makridis, E., Oliva, G., Narahari, K. R., Doostmohammadian, M., Khan, U. A., & Charalambous, T. (2024) · European Control Conference (ECC) · doi:10.23919/ecc64448.2024.10590963 · Cited by 6
“…communicating over a network, each agent j , represented by a node in a graph, has access only to and time-invariant digraphs, Xi et. al. introduced two and ADD-OPT [16] that incorporates gradient tracking Moreover, Nedich et. al. in [17], [18] method, Xi et. al. introduced the Directed-Distributed”
Decentralised Gradient-based Variational Inference for Multi-sensor Fusion and Tracking in Clutter
Li, Q., Gan, R., & Godsill, S. (2024) · International Conference on Information Fusion (FUSION) · doi:10.23919/fusion59988.2024.10706445 · Cited by 2
“In particular, we pro-pose an efficient decentralised natural gradient descent VT (DeNG-VT) algorithm that employs the natural gradient [21] and a gradient tracking strategy [22] to accelerate convergence.”
MethodShahriari-Mehr, F., & Panahi, A. (2024) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2024.3520513 · Cited by 4
“The authors in [23], [30], [31], and [32] combined the push-sum protocol and the gradient tracking techniques and respectively proposed the DEXTRA, Push-DIGing, SONATA, and ADD-OPT algorithms, which are working with fixed step sizes.”
Distributed Center-Based Clustering: A Unified Framework
Armacki, A., Bajović, D., Jakovetić, D., & Kar, S. (2025) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2025.3531292 · Cited by 4
“Another line of work related to ours is that of first-order meth-ods for distributed optimization, e.g., [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58].”
Liu, B., Du, W., & Li, Z. (2025) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2025.3539580 · Cited by 2
“Lemma 3 (See [21]): Under the assumption of undirected topology network, for matrices [38]): Defining M = [ m i,j ] ∈ R 3 × 3 an irreducible positive definite matrix with m ii < (cid:12) , for some (cid:12) > 0 and all i = 1 , 2 , 3 , then ρ ( M ) < if and only if det( (cid:12)I − M ) > 0 .”
Distributed inexact Newton method with adaptive step sizes
Jakovetić, D., Krejić, N., & Malaspina, G. (2025) · Computational Optimization and Applications · doi:10.1007/s10589-025-00666-z · Cited by 5
“We compare DINAS with NN [22], DAN and DAN-LA [42], Newton Tracking [41], DIGing [24] and EXTRA [30].”
InfluentialPerforming accelerated convergence in decentralized economic dispatch over dynamic directed networks
Lv, Y., Xiong, H., Zhang, F., et al. (2025) · Journal of the Franklin Institute · doi:10.1016/j.jfranklin.2025.107611 · Cited by 2
Mobility-aware Decentralized Federated Learning for Autonomous Underwater Vehicles
He, H., Du, J., Jiang, C., Wang, J., & Song, J. (2024) · Globecom · doi:10.1109/globecom52923.2024.10901125 · Cited by 1
“Methods such as gradient tracking [10] and inexact gradient descent [11] have been employed to solve the DFL problem under a time-varying graph, but they fail to take into account the underwater channel and compression problems introduced due to transmission constraints.”
Shi, X., Wan, Y., Wen, G., & Yu, X. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3559905 · Cited by 6
“To achieve linear convergence, several prominent algorithms, including ADMM algorithms [9], DIGing [10] and zero-gradient-sum (ZGS) algorithms [11], [12], [13], [14], [15], are developed for unconstrained DO with strongly convex objectives.”
Adaptive Weighting Push-SUM for Decentralized Optimization With Statistical Diversity
Zhou, Y., Cheng, Y., Xu, L., & Chen, E. (2025) · IEEE Transactions on Control of Network Systems · doi:10.1109/tcns.2025.3566329 · Cited by 7
“Similarly,although gradienttracking[22],[23]isconsideredaneffectivesolutionfor addressingstatisticaldiversity[24],andmanystudies[23],[25], [26]combinedthePush-SUMprotocolwithgradienttracking fordecentralizedoptimization,therearesignificantlimitations.”
On the convergence result of the gradient-push algorithm on directed graphs with constant stepsize
Choi, W., Kim, D., & Yun, S.-B. (2025) · Journal of Global Optimization · doi:10.1007/s10898-025-01506-4 · Cited by 2
“The Push-DIGing/ADD-OPT algorithm [24, 43] can be seen as an extension of the gradient-push algorithm (1.2) with communication of the gradient information.”
InfluentialWang, D. (2025) · IEEE Control Systems Letters · doi:10.1109/lcsys.2025.3576667 · Cited by 2
“Leveraging the (sub)gradient-descent method and the consensus theory, numerous distributed optimization algorithms (control strategies) have been developed in the literature both in discrete-time and continuous-time settings (see [3], [4], [5], [6], [7] and references therein).”
Metadata
- Authors
- Nedić, A., Olshevsky, A., & Shi, W.
- Affiliations
- Boston University (US)
- Arizona State University (US)
- Venue
- SIAM Journal on Optimization
- Published
- 2017-01
- Volume
- 27
- Issue
- 4
- Pages
- 2597-2633
- ISSN
- 1052-6234 (ISSN-L)
- 1095-7189
- OpenAlex
- W2963649943
- Cited by
- 1,069
- References
- 38
- Citations by year
- 2026: 95 · 2025: 147 · 2024: 130 · 2023: 124 · 2022: 118 · 2021: 114 · 2020: 138 · 2019: 106 · 2018: 64 · 2017: 29
- FWCI
- 46.51
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- SIAM Journal on Optimization (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- U.S. Air Force (AF FA95501510394)
- Office of Naval Research (N00014-12-1-0998)
- National Science Foundation (CNS 15-44953)
- Topics
- Physical Sciences › Engineering › Control and Systems Engineering › Distributed Control Multi-Agent Systems
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Optimization and Variational Analysis
- Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
- Keywords
- distributed optimization, dynamic graphs, geometric convergence, R-linear convergence, gradient tracking, strong convexity, fixed step size, undirected graphs, directed graphs
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash and gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.