Abstract
We consider a distributed multi-agent network system where the goal is to minimize a sum of convex objective functions of the agents subject to a common convex constraint set. Each agent maintains an iterate sequence and communicates the iterates to its neighbors. Then, each agent combines weighted averages of the received iterates with its own iterate, and adjusts the iterate by using subgradient information (known with stochastic errors) of its own function and by projecting onto the constraint set.
The goal of this paper is to explore the effects of stochastic subgradient errors on the convergence of the algorithm. We first consider the behavior of the algorithm in mean, and then the convergence with probability 1 and in mean square. We consider general stochastic errors that have uniformly bounded second moments and obtain bounds on the limiting performance of the algorithm in mean for diminishing and non-diminishing stepsizes. When the means of the errors diminish, we prove that there is mean consensus between the agents and mean convergence to the optimum function value for diminishing stepsizes. When the mean errors diminish sufficiently fast, we strengthen the results to consensus and convergence of the iterates to an optimal solution with probability 1 and in mean square.
Similar content being viewed by others
References
Nedić, A., Bertsekas, D.P.: Incremental subgradient method for nondifferentiable optimization. SIAM J. Optim. 12, 109–138 (2001)
Nedić, A., Bertsekas, D.P.: Convergence rate of incremental algorithms. In: Uryasev, S., Pardalos, P.M. (eds.) Stochastic Optimization: Algorithms and Applications, pp. 223–264. Kluwer Academic, Dordrecht (2001)
Pardalos, P.M., Resende, M.G.C.: Handbook of Applied Optimization. Oxford University Press, New York (2002)
Nedić, A., Ozdaglar, A.: Distributed subgradient methods for multi-agent optimization. Trans. Automat. Control 54, 48–61 (2009)
Nedić, A., Ozdaglar, A.: On the rate of convergence of distributed asynchronous subgradient methods for multi-agent optimization. In: Proceedings of the 46th IEEE Conference on Decision and Control, pp. 4711–4716 (2007)
Sundhar, R.S., Nedić, A., Veeravalli, V.V.: Incremental stochastic subgradient algorithms for convex optimization. SIAM J. Optim. 20, 691–717 (2009)
Nedić, A., Ozdaglar, A., Parrilo, P.A.: Constrained consensus and optimization in multi-agent networks. Lab. for Information and Decision Systems Technical Report 2779, Massachusetts Institute of Technology (2008)
Rabbat, M.G., Nowak, R.D.: Distributed optimization in sensor networks. In: Proceedings of International Symposium on Information Processing in Sensor Networks, pp. 20–27 (2004)
Blatt, D., Hero, A.O., Gauchman, H.: A convergent incremental gradient method with constant stepsize. SIAM J. Optim. 18, 29–51 (2007)
Johansson, B., Rabi, M., Johansson, M.: A simple peer-to-peer algorithm for distributed optimization in sensor networks. In: Proceedings of the 46th IEEE Conference on Decision and Control, pp. 4705–4710 (2007)
Johansson, B.: On distributed optimization in networked systems. Ph.D. thesis, Royal Institute of Technology, Stockholm (2008)
Sundhar, R.S.: Distributed optimization in multi-agent systems with applications to distributed regression. Ph.D. thesis, University of Illinois at Urbana-Champaign (2009)
Nedić, A., Ozdaglar, A.: Cooperative distributed multi-agent optimization. In: Daniel, P.P., Eldar, Y.C. (eds.) Convex Optimization in Signal Processing and Communications, pp. 340–386. Cambridge University Press, Cambridge (2010)
Rabbat, M.G., Nowak, R.D.: Quantized incremental algorithms for distributed optimization. IEEE J. Sel. Areas Commun. 23, 798–808 (2005)
Gaivoronski, A.A.: Convergence properties of backpropagation for neural nets via theory of stochastic gradient methods. Optim. Methods Softw. 4, 117–134 (1994)
Solodov, M.V., Zavriev, S.K.: Error stability properties of generalized gradient-type algorithms. J. Optim. Theory Appl. 98, 663–680 (1998)
Bertsekas, D.P., Tsitsiklis, J.N.: Gradient convergence in gradient methods with errors. SIAM J. Optim. 10, 627–642 (2000)
Kiwiel, K.C.: Convergence of approximate and incremental subgradient methods for convex optimization. SIAM J. Optim. 14, 807–840 (2003)
Nedić, A., Bertsekas, D.P.: The effect of deterministic noise in sub-gradient methods. Math. Program. (2009). doi:10.1007/s10107-008-0262-5
Tsitsiklis, J.N.: Problems in decentralized decision making and computation. Ph.D. thesis, Massachusetts Institute of Technology (1984)
Tsitsiklis, J.N., Bertsekas, D.P., Athans, M.: Distributed asynchronous deterministic and stochastic gradient optimization algorithms. IEEE Trans. Automat. Control 31, 803–812 (1986)
Bertsekas, D.P., Tsitsiklis, J.N.: Parallel and Distributed Computation: Numerical Methods. Athena Scientific, Belmont (1997)
Jadbabaie, A., Lin, J., Morse, S.: Coordination of groups of mobile autonomous agents using nearest neighbor rules. IEEE Trans. Automat. Control 48, 998–1001 (2003)
Spanos, D.S., Olfati-Saber, R., Murray, R.M.: Approximate distributed Kalman filtering in sensor networks with quantifiable performance. In: Proceedings of IEEE International Conference on Information Processing in Sensor Networks, pp. 133–139 (2005)
Kar, S., Moura, J.M.F.: Distributed consensus algorithms in sensor networks: link and channel noise. IEEE Trans. Signal Process. 57, 355–369 (2009)
Xiao, L., Boyd, S., Kim, S.-J.: Distributed average consensus with least mean square deviation. J. Parallel Distrib. Comput. 67, 33–46 (2007)
Lobel, I., Ozdaglar, A.: Distributed subgradient methods over random networks. Lab. for Information and Decision Systems Technical Report 2800, Massachusetts Institute of Technology (2008)
Olshevsky, A., Tsitsiklis, J.N.: Convergence speed in distributed consensus and averaging. SIAM J. Control Optim. 48, 33–55 (2009)
Nedić, A., Olshevsky, A., Ozdaglar, A., Tsitsiklis, J.N.: Distributed subgradient algorithms and quantization effects. In: Proceedings of the 47th IEEE Conference on Decision and Control, pp. 4177–4184 (2008)
Huang, M., Manton, J.H.: Stochastic approximation for consensus seeking: mean square and almost sure convergence. In: Proceedings of the 46th IEEE Conference on Decision and Control, pp. 306–311 (2007)
Ermoliev, Y.: Stochastic Programming Methods. Nauka, Moscow (1976)
Ermoliev, Y.: Stochastic quasi-gradient methods and their application to system optimization. Stochastics 9, 1–36 (1983)
Ermoliev, Y.: Stochastic quazigradient methods. In: Ermoliev, Y., Wets, R.J.-B. (eds.) Numerical Techniques for Stochastic Optimization, pp. 141–186. Springer, New York (1988)
Sundhar, R.S., Veeravalli, V.V., Nedić, A.: Distributed and non-autonomous power control through distributed convex optimization. In: Proceedings of the 28th IEEE Conference on Computer Communications INFOCOM, pp. 3001–3005 (2009)
Bertsekas, D.P., Nedić, A., Ozdaglar, A.: Convex Analysis and Optimization. Athena Scientific, Belmont (2003)
Rockafellar, R.T.: Convex Analysis. Princeton University Press, Princeton (1970)
Polyak, B.T.: Introduction to Optimization. Optimization Software, New York (1987)
Billingsley, P.: Probability and Measure. Wiley, New York (1979)
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by P.M. Pardalos.
This work is supported by NSF Career Grant CMMI 07-42538.
Rights and permissions
About this article
Cite this article
Sundhar Ram, S., Nedić, A. & Veeravalli, V.V. Distributed Stochastic Subgradient Projection Algorithms for Convex Optimization. J Optim Theory Appl 147, 516–545 (2010). https://doi.org/10.1007/s10957-010-9737-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10957-010-9737-7