Article Preview
Top1. Introduction
A misconception as to hospital model development is to replace the entire wired backbone with the wireless architecture. This opinion is not realistic. Since the radio signal strength varies constantly with the surrounding environment, such as distance, source of interference (interior and exterior), and the protocol being used etc., its performance is not as stable as the wired networks. Literatures also reveal that Quality of Service (QoS) is not supported in current mainstream Wireless Local Area Network (WLAN) protocols (namely, 802.11a (Latkoski, Hadzi-Velkov, & Popovski, 2005), 802.11b (Banitsas, Song & Owens, 2004) and 802.11g (Wijesinha, Song, Krishnan, Mathur, Ahn, & Shyamasundar, 2005)). Although 802.11e (Xiao, 2004). is introduced as an emerging standard to provision QoS features on a contention-based channel access scheme, extensive testing is still required for the deployment in medicine areas. Hence, a more integrated approach is expected to incorporate wireless technology into the existing backbone network, ensuring that the centralised service are maintained by wired architecture, while cable free flexibility and escalated bandwidth still satisfy the peripheral mobile service as well as emerging medical services (e.g. heart monitor and telemetry).
A number of researchers (Varshney, 2007) have made effort in designing an intellectual Emergency Unit (EU) model that extends accessibility and interoperability against the legacy system, yet a generic hospital model is not fully developed. The complexity is due to the characteristics of hospital scenarios. Unlike the commercial and educational networks (Fang, Hong, Yu-heng, & Li 2007), a standard hospital normally involves life-critical application, multimedia application and office-oriented application, and therefore the network model requires more specific performance metrics to evaluate data, audio and video traffic under different conditions. Especially, in the emergency units, timely response to the biomedical signals or alarm is highly critical to saving lives, so mean packet delay and packet loss are sensitive at this point. As the network scenario switches to radiology room or intensive care unites, the tolerated range for mean packet delay and packet loss must be varied. The patterns of transmitting data, audio and video traffic are fully specified in Chigan and Oberoi (2006). Researchers can observe the network behaviours by means of simulation tools for data validity and verification purposes.
Out of the numerous design and modelling issues, the main ones encountered network practitioners include validity and feasibility of network architecture, network protocol and the strategies that figure out performance metrics, user capacity, and propagation environment. Presumably, the network architecture offers fast and reliable connectivity for hospital-wide applications. However, different traffic demands on different security and quality-based provision. Explicitly, voice and video traffic requires a robust gateway to maintain authentication and sufficient bandwidth for each session, while data traffic tends to more rely on backup, security policy and Virtual Private Network (VPN) technology dedicated for the remote users. Therefore, a preliminary feasibility research is required for deploying such technologies in the hospital scenarios. The network architecture is not validated until cost, availability, scalability and all performance data have met the predefined standards. Likewise, the prospective WLAN protocols (802.11a and 802.11g) need to be fully justified in terms of throughput, mean packet delay and packet drop, etc. This research paper aims to develop a generic hospital network model and evaluate its performance for the integrated network scenarios, including radiology room, and Emergency (A & E). This research makes use of computer simulation and discusses various aspects of the network design, so as to discover the performance behaviour pertaining to effect of traffic type, traffic load and network size. Intuitive insight is also sought to provide guidelines for novice users in developing hospital simulation models.