An opportunistic cross-layer architecture for voice in multi-hop wireless LANs

Suhaib A. Obeidat, Violet Syrotiuk

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


We propose an opportunistic cross-layer architecture for adaptive support of Voice over IP in multi-hop wireless LANs. As opposed to providing high call quality, we target emergencies where it is important to communicate, even if at low quality, no matter the harshness of the network conditions. With the importance of delay on voice quality in mind, we select adaptation parameters that control the ratio of real-time traffic load to available bandwidth. This is achieved in two ways: minimizing the load and maximizing the bandwidth. The PHY/MAC interaction improves the use of the spectral resources by opportunistically exploiting rate-control and packet bursts, while the MAC/application interaction controls the demand per source through voice compression. The objective is to maximize the number of calls admitted that satisfy the end-to-end delay budget. The performance of the protocol is studied extensively in the ns-2 network simulator. Results indicate that call quality degrades as load increases and overlonger paths, and a larger packet size improves performance. For long paths having low-quality channels, forward error correction, header compression, and relaxing the delay budget of the system are required to maintain call admission and quality. The proposed adaptive protocol achieves high performance improvements over the traditional, non-adaptive approach.

Original languageEnglish (US)
Pages (from-to)419-439
Number of pages21
JournalInternational Journal of Communication Systems
Issue number4
StatePublished - Apr 2009


  • Cross-layer
  • Opportunistic
  • QoS
  • Simulation

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Electrical and Electronic Engineering


Dive into the research topics of 'An opportunistic cross-layer architecture for voice in multi-hop wireless LANs'. Together they form a unique fingerprint.

Cite this