Intersession Network Coding for Two Simple Multicast Sessions

In this paper, the problem of transmitting two symbols from two sources to multiple destinations using network coding is considered for general directed acyclic graphs. When all destination nodes are interested in both symbols, the feasibility of this problem is characterized by the classic min-cut maxflow theorem. This work focuses on the scenarios in which each destination is interested in one source symbol. If no coding is allowed at intermediate nodes, the feasibility of this traditional two-session routing problem is equivalent to the existence of two edge-disjoint sets of paths. Similarly, in this paper, it is proven that the feasibility of this inter-session network coding problem is equivalent to the existence of multiple sets of paths with controlled edge overlaps. In particular, for the degenerate case of two sources and two destinations, it is proven that a binary field is sufficient and an optimal network coding solution must be of one of the three different forms, which includes the well-studied butterfly graph as a special case.

[1]  Peter Sanders,et al.  Polynomial time algorithms for network information flow , 2003, SPAA '03.

[2]  Ness B. Shroff,et al.  Beyond the Butterfly - A Graph-Theoretic Characterization of the Feasibility of Network Coding with Two Simple Unicast Sessions , 2007, 2007 IEEE International Symposium on Information Theory.

[3]  R. Koetter,et al.  Network coding from a network flow perspective , 2003, IEEE International Symposium on Information Theory, 2003. Proceedings..

[4]  Serap A. Savari,et al.  Edge-Cut Bounds on Network Coding Rates , 2006, Journal of Network and Systems Management.

[5]  Randall Dougherty,et al.  Insufficiency of linear coding in network information flow , 2005, IEEE Transactions on Information Theory.

[6]  Zongpeng Li,et al.  Network Coding in Undirected Networks , 2004 .

[7]  Yunnan Wu,et al.  On Constructive Multi-Source Network Coding , 2006, 2006 IEEE International Symposium on Information Theory.

[8]  John E. Hopcroft,et al.  The Directed Subgraph Homeomorphism Problem , 1978, Theor. Comput. Sci..

[9]  Zhen Zhang,et al.  An outer bound for multisource multisink network coding with minimum cost consideration , 2006, TNET.

[10]  Ness B. Shroff,et al.  Optimization Based Rate Control for Communication Networks with Inter-Session Network Coding , 2008, IEEE INFOCOM 2008 - The 27th Conference on Computer Communications.

[11]  Lihua Song,et al.  Zero-error network coding for acyclic network , 2003, IEEE Trans. Inf. Theory.

[12]  Robert D. Kleinberg,et al.  On the capacity of information networks , 2006, IEEE Transactions on Information Theory.

[13]  R. Koetter,et al.  An algebraic approach to network coding , 2001, Proceedings. 2001 IEEE International Symposium on Information Theory (IEEE Cat. No.01CH37252).

[14]  Rudolf Ahlswede,et al.  Network information flow , 2000, IEEE Trans. Inf. Theory.

[15]  Shuo-Yen Robert Li,et al.  Linear network coding , 2003, IEEE Trans. Inf. Theory.

[16]  Baochun Li,et al.  Network Coding : The Case of Multiple Unicast Sessions , 2004 .

[17]  Vijay V. Vazirani,et al.  On the capacity of multiple unicast sessions in undirected graphs , 2005, ISIT.

[18]  Muriel Médard,et al.  Network Coding for Multiple Unicasts: An Approach based on Linear Optimization , 2006, 2006 IEEE International Symposium on Information Theory.

[19]  Sae-Young Chung,et al.  Network coding , 2008, Journal of Communications and Networks.

[20]  Christina Fragouli,et al.  Information flow decomposition for network coding , 2006, IEEE Transactions on Information Theory.