Distributed aggregative optimization with quantized communication

Ziqin Chen; Shu Liang

Kybernetika (2022)

  • Volume: 58, Issue: 1, page 123-144
  • ISSN: 0023-5954

Abstract

top
In this paper, we focus on an aggregative optimization problem under the communication bottleneck. The aggregative optimization is to minimize the sum of local cost functions. Each cost function depends on not only local state variables but also the sum of functions of global state variables. The goal is to solve the aggregative optimization problem through distributed computation and local efficient communication over a network of agents without a central coordinator. Using the variable tracking method to seek the global state variables and the quantization scheme to reduce the communication cost spent in the optimization process, we develop a novel distributed quantized algorithm, called D-QAGT, to track the optimal variables with finite bits communication. Although quantization may lose transmitting information, our algorithm can still achieve the exact optimal solution with linear convergence rate. Simulation experiments on an optimal placement problem is carried out to verify the correctness of the theoretical results.

How to cite

top

Chen, Ziqin, and Liang, Shu. "Distributed aggregative optimization with quantized communication." Kybernetika 58.1 (2022): 123-144. <http://eudml.org/doc/297603>.

@article{Chen2022,
abstract = {In this paper, we focus on an aggregative optimization problem under the communication bottleneck. The aggregative optimization is to minimize the sum of local cost functions. Each cost function depends on not only local state variables but also the sum of functions of global state variables. The goal is to solve the aggregative optimization problem through distributed computation and local efficient communication over a network of agents without a central coordinator. Using the variable tracking method to seek the global state variables and the quantization scheme to reduce the communication cost spent in the optimization process, we develop a novel distributed quantized algorithm, called D-QAGT, to track the optimal variables with finite bits communication. Although quantization may lose transmitting information, our algorithm can still achieve the exact optimal solution with linear convergence rate. Simulation experiments on an optimal placement problem is carried out to verify the correctness of the theoretical results.},
author = {Chen, Ziqin, Liang, Shu},
journal = {Kybernetika},
keywords = {distributed aggregative optimization; multi-agent network; quantized communication; linear convergence rate},
language = {eng},
number = {1},
pages = {123-144},
publisher = {Institute of Information Theory and Automation AS CR},
title = {Distributed aggregative optimization with quantized communication},
url = {http://eudml.org/doc/297603},
volume = {58},
year = {2022},
}

TY - JOUR
AU - Chen, Ziqin
AU - Liang, Shu
TI - Distributed aggregative optimization with quantized communication
JO - Kybernetika
PY - 2022
PB - Institute of Information Theory and Automation AS CR
VL - 58
IS - 1
SP - 123
EP - 144
AB - In this paper, we focus on an aggregative optimization problem under the communication bottleneck. The aggregative optimization is to minimize the sum of local cost functions. Each cost function depends on not only local state variables but also the sum of functions of global state variables. The goal is to solve the aggregative optimization problem through distributed computation and local efficient communication over a network of agents without a central coordinator. Using the variable tracking method to seek the global state variables and the quantization scheme to reduce the communication cost spent in the optimization process, we develop a novel distributed quantized algorithm, called D-QAGT, to track the optimal variables with finite bits communication. Although quantization may lose transmitting information, our algorithm can still achieve the exact optimal solution with linear convergence rate. Simulation experiments on an optimal placement problem is carried out to verify the correctness of the theoretical results.
LA - eng
KW - distributed aggregative optimization; multi-agent network; quantized communication; linear convergence rate
UR - http://eudml.org/doc/297603
ER -

References

top
  1. Barbarossa, S., Sardellitti, S., Lorenzo, P. D., , IEEE Signal Process. Mag. 31 (2014), 45-55. DOI
  2. Cao, X., Liu, K. J. R., , IEEE Trans. Automat. Control 66 (2021), 1278-1285. MR4226775DOI
  3. Cheng, S., Liang, S., , Kybernetika 56 (2020), 559-577. MR4131743DOI
  4. Chen, J., Sayed, A., , IEEE Trans. Signal Process. 60 (2012), 4289-4305. MR2960496DOI
  5. Deng, Z., Liang, S., , Automatica 99 (2019), 246-252. MR3876174DOI
  6. Persis, C. De, Grammatico, S., , IEEE Trans. Automat. Control 65 (2020), 2171-2176. MR4091832DOI
  7. Horn, R. A., Johnson, C. R, Matrix Analysis., Cambridge University Press, New York 2012. Zbl0801.15001MR2978290
  8. Jakovetic, D., Moura, J. M. F., Xavier, J., , IEEE Trans. Automat. Control 60 (2015), 922-936. MR3340785DOI
  9. Kajiyama, Y., Hayashi, N., Takai, S., , IEEE Trans. Automat. Control 66 (2020), 1254-1261. MR4226772DOI
  10. Lan, G., Lee, S., Zhou, Y., , Math. Program. 180 (2020), 237-284. MR4062837DOI
  11. Li, P., Hu, J., , Control Theory Technol. 19 (2021), 499-506. MR4356235DOI
  12. Li, P., Hu, J., Qiu, L., Zhao, Y., Ghosh, B., , IEEE Trans. Control Netw. Syst., Early Access (2021). DOI
  13. Li, X., Xie, L., Hong, Y., , Int. J. Robust Nonlinear Control 29 (2019), 3252-3266. MR3973593DOI
  14. Li, X., Xie, L., Hong, Y., , IEEE Trans. Automat. Control. Early Access (2021). DOI
  15. Ma, J., Yu, X., Liu, L., Feng, G., , IEEE Trans. Control Netw. Syst., Early Access (2021). DOI
  16. Magnussson, S., Shokri-Ghadikolaei, H., Li, N., , IEEE Trans. Signal Process. 68 (2020), 6101-6116. MR4177712DOI
  17. Msechu, E. J., Giannakis, G. B., , IEEE Trans. Signal Process. 60 (2011), 400-414. MR2932127DOI
  18. Nedic, A., 10.1109/MSP.2020.2975210, IEEE Signal Process. Mag. 73 (2020), 92-101. DOI10.1109/MSP.2020.2975210
  19. Nedic, A., Ozdaglar, A., , IEEE Trans. Autom. Control 54 (2009), 48-61. MR2478070DOI
  20. Pu, Y., Zeilinger, M. N., Jones, C. N., , IEEE Trans. Automat. Control 62 (2016), 2107-2120. MR3641434DOI
  21. Ren, W., Beard, R. W., Distributed consensus in multi-vehicle cooperative control., Springer, London 2008. 
  22. Shi, W., Ling, Q., Wu, G., Yin, W., , SIAM J. Optim. 25 (2015), 944-966. MR3343366DOI
  23. Tardos, E., Vazirani, V. V., , Algorithmic Game Theory (2007), 3-28. MR2391748DOI
  24. Wang, X., Deng, Z., Ma, S., Xian, D., , Kybernetika 53 (2017), 179-194. MR3638563DOI
  25. Wang, Y., Lin, P., Qin, H., , Kybernetika 53 (2017), 595-611. MR3730254DOI
  26. Xu, J., Zhu, S., Soh, Y. C., Xie, L., , IEEE Trans. Automat. Control 63 (2018), 3809-3824. MR3875379DOI
  27. Yi, P., Hong, Y., , IEEE Trans. Control Netw. Syst. 1 (2014), 380-392. MR3303147DOI
  28. Yi, P., Hong, Y., Liu, F., , Automatica 74 (2016), 259-269. MR3569392DOI
  29. Yuan, D., Hong, Y., Ho, D., Jiang, G., , Automatica 90 (2018), 196-203. MR3764399DOI
  30. Zhang, X., Liu, J., Zhu, Z., Bentley, E. S., , In: Proc. IEEE Conf. Comput. Commun. 2019, 2431-2439. DOI
  31. Zhu, S., Hong, M., Chen, B., , In: Proc. IEEE Int. Conf. Acoust., Speech Signal Process 2016, pp. 4134-4138. DOI

Citations in EuDML Documents

top
  1. Chenyang Liu, Xiaohua Dou, Yuan Fan, Songsong Cheng, A penalty ADMM with quantized communication for distributed optimization over multi-agent systems
  2. Xianlin Zeng, Lihua Dou, Jinqiang Cui, Distributed accelerated Nash equilibrium learning for two-subnetwork zero-sum game with bilinear coupling
  3. Yikun Zeng, Yiheng Wei, Shuaiyu Zhou, Dongdong Yue, Distributed optimization via active disturbance rejection control: A nabla fractional design
  4. Zhaoxu Wang, Chao Zhai, Hehong Zhang, Gaoxi Xiao, Guanghou Chen, Yulin Xu, Coordination control and analysis of TCSC devices to protect electrical power systems against disruptive disturbances

NotesEmbed ?

top

You must be logged in to post comments.

To embed these notes on your page include the following JavaScript code on your page where you want the notes to appear.

Only the controls for the widget will be shown in your chosen language. Notes will be shown in their authored language.

Tells the widget how many notes to show per page. You can cycle through additional notes using the next and previous controls.

    
                

                
Note: Best practice suggests putting the JavaScript code just before the closing </body> tag.