Convergence of Consensus Models With Stochastic Disturbances

被引:73
作者
Aysal, Tuncer Can [1 ]
Barner, Kenneth E. [2 ]
机构
[1] Cornell Univ, Sch Elect & Comp Engn, Commun Res Signal Proc Grp, Ithaca, NY 14853 USA
[2] Univ Delaware, Dept Elect & Comp Engn, Signal Proc & Commun Grp, Newark, DE 19716 USA
基金
美国国家科学基金会;
关键词
Convergence in expectation; convergence of random sequences; convergence to consensus; distributed average consensus; gossip algorithms; mean square error; noisy gossip algorithms; nonsum-preserving gossip algorithms; quantized gossip algorithms; sum-preserving gossip algorithms; DISTRIBUTED AVERAGE CONSENSUS; NETWORKS; QUANTIZATION; ALGORITHMS; TOPOLOGY; SIGNALS; DITHER; AGENTS; LINKS;
D O I
10.1109/TIT.2010.2050940
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We consider consensus algorithms in their most general setting and provide conditions under which such algorithms are guaranteed to converge, almost surely, to a consensus. Let {A(t), B(t)} is an element of R(NxN) be (possibly) stochastic, nonstationary matrices and {x(t), m(t)} is an element of R(Nx1) be state and perturbation vectors, respectively. For any consensus algorithm of the form x(t + 1) = A(t)x(t) + B(t)m(t), we provide conditions under which consensus is achieved almost surely, i.e., Pr{lim(t ->infinity) x(t) = c1} = 1 for some c is an element of R. Moreover, we show that this general result subsumes recently reported results for specific consensus algorithms classes, including sum-preserving, nonsum-preserving, quantized, and noisy gossip algorithms. Also provided are the epsilon-converging time for any such converging iterative algorithm, i.e., the earliest time at which the vector x(t) is epsilon close to consensus, and sufficient conditions for convergence in expectation to the average of the initial node measurements. Finally, mean square error bounds of any consensus algorithm of the form discussed above are presented.
引用
收藏
页码:4101 / 4113
页数:13
相关论文
共 38 条
  • [1] AYSAL T, 2008, P IEEE C DEC CONTR C
  • [2] AYSAL T, 2007, P IEEE STAT SIGN PRO
  • [3] Aysal T. C., 2008, P 2008 IEEE INF THEO
  • [4] Aysal T. C., 2007, P ALL C COMM CONTR C
  • [5] Distributed average consensus with dithered quantization
    Aysal, Tuncer Can
    Coates, Mark J.
    Rabbat, Michael G.
    [J]. IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2008, 56 (10) : 4905 - 4918
  • [6] Benezit F., 2007, P ALL C COMM CONTR C
  • [7] Benezit Florence, 2007, P 45 ANN ALL C COMM
  • [8] Bertsekas D.P., 1989, PARALLEL DISTRIBUTED
  • [9] Randomized gossip algorithms
    Boyd, Stephen
    Ghosh, Arpita
    Prabhakar, Balaji
    Shah, Devavrat
    [J]. IEEE TRANSACTIONS ON INFORMATION THEORY, 2006, 52 (06) : 2508 - 2530
  • [10] Dimakis A. G., 2008, IEEE T SIGNAL PROCES, V56