Optimal cross-layer scheduling for multicast in multi-channel wireless networks
Yong Liu · 2009
We consider a generic multi-channel wireless network modeled by a hyper graph G = (V,H), where V is the set of nodes, and H is the set of broadcast links (potentially operating at different channels). Each broadcast link l ∈ H can be represented by a hyper-arc l = 〈i,J〉, with i ∈ V being the transmitter and J ⊆ H the set of intended receivers within i’s broadcast range. To model selective broadcast and variable range/rate broadcast from a transmitter, we allow one node to have multiple hyper-arcs, each of which has a different subset of intended receivers. Due to interference between adjacent transmissions, not all broadcast links can be activated simultaneously. Let ziJ be the transmission rate on link 〈i,J〉, and ˆ Z be the set of rate vectors that can be scheduled at any given time. Through time sharing between different rate vectors in ˆ Z, the feasible link rate region of the whole network can be characterized by the convex hull of ˆ Z, Z � CH ( ˆ Z). For clarity of presentation, we start with a single multicast session consisting of a source s and a set of receivers T ⊆ V. We will study the multiple multicast sessions case in Section 3. For a single multicast session, we are interested in the following questions: 1. What is the highest rate at which the source s can multicast data to all receivers in T? 2. How do we achieve the highest rate through joint data coding, routing, and link scheduling? It has been shown recently that the optimal multicast rate can be achieved through network coding in