MeshChord: A Location-Aware, Cross-Layer Specialization of Chord for Wireless Mesh Networks (concise contribution)

Evaluating load balancing in peer-to-peer resource sharing algorithms for wireless mesh networks

“…The work that is more closely related to this paper is [3], in which we investigated the performance of the Chord P2P algorithm in typical wireless mesh network scenarios through extensive, packet-level simulations. Furthermore, we proposed the MESHCHORD specialization of the basic Chord design, which, by explicitly accounting for peculiar features of mesh networks, reduces total network-level traffic of as much as 40% with respect to Chord.…”

“…Recently, we have proposed MESHCHORD [3], a specialization for mesh networks of the well-known Chord algorithm [15], which exploits i) the availability of a wireless infrastructure and ii) the I-hop broadcast nature of wireless communications, to realize a) location-aware ID assignment to nodes, and b) implementation of a cross-layering technique that bridges the MAC to the middleware layer.…”

“…Yet, important issues are left unaddressed in [3], such as whether the load induced by Chord/MESHCHORD is evenly distributed among the wireless routers composing the mesh network. Load unbalancing is a major cause of performance degradation in mesh networks, as clients connected to relatively highly loaded routers experience a much lower QoS than those connected to relatively lightly loaded ones.…”

“…The second contribution of this paper is considering another method for resolving queries (key lookup operation) with respect to the one considered in [3]. More specifically, while in [3] we considered iterative query resolution, in which the intermediate results of a query are returned to the node which issued the query, in this paper we consider also the potentially more efficient recursive query resolution, in which intermediate results of a query are directly forwarded to the next node in charge of query resolution (details in Section 3).…”

“…More specifically, while in [3] we considered iterative query resolution, in which the intermediate results of a query are returned to the node which issued the query, in this paper we consider also the potentially more efficient recursive query resolution, in which intermediate results of a query are directly forwarded to the next node in charge of query resolution (details in Section 3). Since the query resolution primitive is invoked not only to resolve client-generated queries, but also to proactively maintain the overlay structure, a considerable reduction in packet overhead with respect to the case of iterative query resolution is expected.…”