IP has been tried and tested successfully in fixed networks for data services. It has become a de facto standard for data communications on the Internet by connecting universities, businesses, and individuals all over the world. In this section the critical limitations of IP applicability to wireless networks are discussed, such as; addressing incompatibilities, wireless link characteristics, mobility (roaming and handover), security and Quality of Service (QoS).

Technical overview

IP in wireless networks plays different roles due to the varying capabilities and applications supported in these technologies. For example, Personal Area Networks (PAN) are concerned with short-range communications within a home or a Small Office/Home Office (SOHO) environment for wireless connectivity and for applications like simple file and data transfer. Existing IP protocols with necessary adaptations to specific PAN radio technologies (e.g., Bluetooth, IEEE 802.15) can help achieve those goals.

WLAN, on the other hand, is meant for broadband wireless access in a larger environment, such as airports, hotels, universities, and malls. The intent of WLAN is similar to the Ethernet but provides a link-level wireless solution. Therefore, WLAN, just like Ethernet, can support IP on the network layer and various IP applications that are/have been developed based on the capabilities of the higher-layer protocols over IP.

Current cellular networks (2G and 3G) are some of the most complex network architectures ever deployed on a large scale and provide a complete solution for addressing, security, paging, mobility, accounting, and so on. 2G cellular radios have limited bandwidth and costly radio resources. In 3G cellular, although the bandwidth is higher, the radio resources are still expensive. Cellular networks make use of strict radio admission control to allocate radio resources based on demand and availability. The IP suite of protocols is not well optimized for low-bandwidth and limited resource links.

Cellular networks provide idle mobility, also called roaming, and active call handoff functions to users without any noticeable glitch to allow for seamless mobility within an operator network and across different operator networks. All these features have been tried and tested and are currently in use on a large scale. The IETF is presently defining mobility solutions with IP (versions 4 and 6) that include idle mobility and handover. The IETF also formed a working group to handle internetwork roaming for Authentication, Authorization, and Accounting (AAA) functions for large-scale deployments.

Recently introduced 3G cellular networks provide higher bandwidth in their packet core network and offer multimedia applications and Web services as a first step of service integration to the Internet. These IP applications are mainly data centric and work complementary to existing cellular voice services by making use of existing Internet infrastructure support and protocols. But they are not suitable to replace cellular functionality without due enhancements.

The first enhancements include support for real-time call establishments over IP networks leading to the development of an IP-based call/session control signalling (Session Initiation Protocol, SIP) to handle calls and other multimedia sessions.

Cellular networks exclusively use the SS7 (Signalling System #7) backbone in the core network to carry signalling. SS7 offers a robust and highly reliable network to carry signalling messages. Cellular networks have been designed to connect seamlessly to the PSTN, which provides landline telecommunication services. This enables mobile users to make calls to landline phones and vice versa. To remain connected to the PSTN network, IP networks must make use of gateways that will perform signalling and media interworking on the border of IP networks and PSTN.

In order for IP to be applied ubiquitously to wireless networks, it must overcome the wireless challenges.

Part of the thesis: Wireless IP, The Killer Application !?