ABSTRACT
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
The availability of multiple data sources provides extra potential to improve wireless
multimedia service quality. However, the exploration of such multisource selection
has largely been ignored in literature. In this chapter we present a new solution to
optimize the receiver-side multimedia quality by coordinating the transmission of
multiple media sources, using a lossy wireless peer-to-peer (P2P) network an an ex-
ample scenario, while assuring the latency constraint. The major advantages of the
coordinative multisource selection solution are twofold. First, it distributes workload
to each media source by optimal grouping of multimedia frames on each peer host.
Second, optimal channel coding rates are allocated to multimedia frames transmitted
on each path. To reduce the computing complexity, the global optimal solution for
all the multimedia frames is divided into multiple local optimal solutions. Specif-
ically, (1) we divide all transmission paths into two groups in light of related bit
error rates and path-pass probability, and allocate the same channel coding rate to
each group, and we (2) find the optimal data source for each multimedia frame ac-
cording to the value of effective transmission capacity (ETC). The simulation results
show that the simplified strategy works as good as the global optimal solution, and
it significantly improves the end-received multimedia quality under different latency
constraints. This work casts new insights to future wireless multimedia streaming
solution provision, by exploring the potential of coordinative multisource selection.
With recent technological advancements of broadband radio frequency access, mul-
timedia information delivery over modern wireless communication networks (such
as P2P networks [1, 2]) demonstrates considerable potential for a wide variety of
applications, such as IPTV [3], video on demand (VoD) [4], file and media con-
tent sharing [5], etc. In such systems, media flows transmitted from multiple sources
can be jointly decoded at the receiver side, providing considerable source diversity
gains. Compared with traditional data P2P networks, wireless multimedia P2P net-
works have two major unique characteristics: stringent multimedia service quality
requirement and real-time streaming latency requirement. Thus, how to provide qual-
ity multimedia service with bounded latency performance becomes the key research
challenge.
