Back to Search Start Over

Hypergraph-based data link layer scheduling for reliable packet delivery in wireless sensing and control networks with end-to-end delay constraints

Authors :
Kam-Yiu Lam
Edward Chan
Song Han
Qingchun Chen
Deji Chen
Mark Nixon
Mao Yan
Pingzhi Fan
Source :
Information Sciences. 278:34-55
Publication Year :
2014
Publisher :
Elsevier BV, 2014.

Abstract

Many mission-critical and safety-critical applications in networked wireless sensing and control systems have stringent reliability requirements and timing constraints on end-to-end ( E2E ) packet delivery. Late arrivals of packets could severely degrade overall system performance and cause serious problems in system operation. In this paper, we study the data link layer scheduling problem to maximize the reliability of E2E packet delivery in TDMA-based wireless sensing and control networks ( WSCNs ) subject to specified delay constraints. We propose to organize the physical network nodes into logical hypernodes and form a hypergraph for improved scheduling flexibility. Based on the hypergraph, we introduce two data link layer scheduling schemes to maximize the E2E reliability in packet delivery without violating any delay constraints of the packets. The first scheme, named dedicated scheduling , decides how many time slots (TSs) for each hypernode along the path to the destination should be allocated to transmit a packet, and the packets are only transmitted in their scheduled TSs. The second scheme, named shared scheduling , allows the packets to share their scheduled TSs, and thus further improve the E2E reliability. We apply these two schemes in both single-path routing and any-path routing in WSCNs. Asymptotic analysis of the proposed schemes is provided and extensive simulation experiments are performed to illustrate their effectiveness in improving the E2E reliability of packet delivery under different network settings.

Details

ISSN :
00200255
Volume :
278
Database :
OpenAIRE
Journal :
Information Sciences
Accession number :
edsair.doi...........33e27ee844d0e0866d8a84ce2b4b475b