Razvan Cristescu

Title: 
Trees for correlated data gathering

Abstract: 
This talk will address the problem of correlated data gathering by a
network with a sink node and a tree communication structure, where the
goal is to minimize the total transmission cost of transporting the
information collected by the nodes, to the sink node. I will analyze
two lossless coding strategies: a joint entropy coding model with
explicit communication where coding is simple and transmission
optimization is difficult, and a Slepian-Wolf model where optimal
coding is complex and transmission optimization is simple. For the
explicit communication case a reduction from the min-set cover problem
shows that building an optimal data gathering tree is NP-hard. We
present a constant approximation algorithm based on the SLT
(shallow-light tree), and a heuristic based on a combination of the
SPT (shortest path tree) and k-TSP (multiple traveling salesmen). For
the Slepian-Wolf setting, the optimal data gathering tree is the SPT
and we derive a closed-form solution for the rate allocation. The
structure of the solution suggests an efficient distributed
approximation algorithm. 

References: 

S. Khuller, B. Raghavachari and N. Young. Balancing Minimum Spanning
and Shortest Path Trees, SODA 1993, pp. 243-250. 

R. Cristescu, B. Beferull-Lozano, M. Vetterli and R. Wattenhofer,
Network Correlated Data Gathering with Explicit Communication:
NP-Completeness and Algorithms, to appear in IEEE/ACM Transactions on
Networking, April 2006.