Efficient network codes for cyclic networks

In this work we address the problem of network codes for cyclic networks. We show that network codes can be constructed for cyclic networks as long as at least one edge in each cycle has a delay, but it is not required that every edge would have a delay. We then present the algorithm for constructing an optimal multicast network code, developed in our previous work, and analyze its computational complexity, showing that it is polynomial in the graph size. We discuss the properties of the resulting codes, and show the ability to modify the code in a localized manner when sinks are added or removed. This property is also applicable to acyclic networks. Finally, we propose the sequential decoding algorithm we developed in an earlier work for decoding the resulting codes. For this we analyze its decoding delay, for both acyclic and cyclic networks

[1]  Meir Feder,et al.  Convolutional Network Codes for Cyclic Networks , 2005 .

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

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

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

[5]  Lihua Song,et al.  A separation theorem for single-source network coding , 2006, IEEE Transactions on Information Theory.

[6]  Joseph Naor,et al.  Approximating Minimum Feedback Sets and Multicuts in Directed Graphs , 1998, Algorithmica.

[7]  Meir Feder,et al.  Convolutional network codes , 2004, International Symposium onInformation Theory, 2004. ISIT 2004. Proceedings..

[8]  Peter Sanders,et al.  Polynomial time algorithms for multicast network code construction , 2005, IEEE Transactions on Information Theory.