DocumentCode :
1414740
Title :
Scheduling for Network-Coded Multicast
Author :
Traskov, Danail ; Heindlmaier, Michael ; Médard, Muriel ; Koetter, Ralf
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Massachusetts Inst. of Technol., Cambridge, MA, USA
Volume :
20
Issue :
5
fYear :
2012
Firstpage :
1479
Lastpage :
1488
Abstract :
We consider multicasting using random linear network coding over a multihop wireless network in the bandwidth limited regime. We address the associated medium access problem and propose a scheduling technique that activates hyperarcs rather than links, as in classical scheduling approaches. We encapsulate the constraints on valid network configurations in a conflict graph model and formulate a joint optimization problem taking into account both the network coding subgraph and the schedule. Next, using Lagrangian relaxation, we decompose the overall problem into two subproblems, a multiple-shortest-paths problem and a maximum weighted stable set (MWSS) problem. We show that if we use a greedy heuristic for the MWSS part of the problem, the overall algorithm is completely distributed. We provide extensive simulation results for both the centralized optimal and the decentralized algorithms. The optimal algorithm improves performance by up to a factor of two over widely used techniques such as orthogonal or two-hop-constrained scheduling. The decentralized algorithm is shown to buy its distributed operation with some throughput losses. Experimental results on randomly generated networks suggest that these losses are not large. Finally, we study the power consumption of our scheme and quantify the tradeoff between power and bandwidth efficiency.
Keywords :
access protocols; linear codes; multicast communication; network coding; network theory (graphs); optimisation; radio networks; random codes; scheduling; set theory; Lagrangian relaxation; MWSS; bandwidth limited regime; centralized optimal algorithm; classical scheduling approach; conflict graph model; decentralized algorithm; greedy heuristic; hyperarc; joint optimization problem; maximum weighted stable set; medium access problem; multihop wireless network; multiple shortest path problem; network coded multicasting; network coding subgraph; network configuration; random linear network coding; randomly generated network; throughput loss; Interference; Network coding; Optimization; Scheduling; Vectors; Wireless networks; Multicast; network coding; scheduling; wireless networks;
fLanguage :
English
Journal_Title :
Networking, IEEE/ACM Transactions on
Publisher :
ieee
ISSN :
1063-6692
Type :
jour
DOI :
10.1109/TNET.2011.2180736
Filename :
6122470
Link To Document :
بازگشت