Communications mobility is the ability for a device to communicate with another device from various locations over one or more access links (for example, Ethernet, 802.11, and so on). The ultimate goalfull mobilityadds the requirement of maintaining communications across access link changes. Just as you don't notice when your cell phone changes cell towers, the end user should never notice access link changes. In this context, mobility can mean changing from one subnet to another on an Ethernet network or making a telephone call in the car while passing from one cell site to another. Mobility can even mean moving from a high-speed Ethernet network at an office desk to a WLAN connection in a conference room. Regardless, all mobility solutions must be capable of locating a device and delivering traffic to that device. Specifically, a mobility solution must address the following four requirements:
Every communications mobility protocol can address these requirements in a different way. For example, some handle as much as possible in the core of the network, others at the edge of the network.
For a network to deliver traffic to an end device, the network must know where to find the device. Networks use the following two approaches to find a device in a network:
Even though these are two completely different systems, the end result is the sametwo parties in an airport can communicate with each other. As explored in Chapter 2, "Understanding Mobile IP," Mobile IP takes a similar approach.
Location discovery, whether it's reactive or proactive, is the method by which the network knows how to deliver traffic to the mobile device. The specific implementation of the location discovery is closely tied to the underlying technology and predicted traffic model.
While the term move detection has some clearly physical connotations, in this context, it is almost always logical. To put it simply, move detection is a change in available access links. Unlike fixed networking, where most devices have only one connection into the network, mobile devices often have several different paths available at a given time. The mobility protocol is then responsible for determining which of these paths is best and communicating this path as the device's current location. Cellular networks, for example, often have an overlap in the coverage area of their cell sites. The network then makes a determination, based on signal quality and other factors, of which cell site the device should associate with. In most cases, a logical move is the result of a physical move; however, if an access link fails, it is also treated as movement.
For protocols that provide full mobility, sessions can be maintained when a device changes its access link. A location change that's based on move detection is called a handover.
Each time movement is detected, the system must determine whether this movement is equivalent to a change in location. Some protocols take a proactive approach, evaluating the optimal location with each movement. Others are reactive, looking for a new optimal link only when the existing link is no longer available. In some cases, the protocol is designed to allow the user to customize his handover algorithms.
Both update signaling and path (re)establishment are concepts that already exist in IP routing protocols. Update signaling is equivalent to a routing update. Update signaling communicates the current location of the mobile device to the rest of the network. Update signaling communicates location changes and can inform the network that a device is still active in its current location. Signals can be sent on a per-device basis or they can be aggregated.
Path (re)establishment is similar to the concept of convergence in IP routing protocols. When a device signals that it has changed its location, the network must have some way to establish a path to the device, or if a path was already in use, to change that path so that traffic can be delivered to the device at its new location. This can be accomplished in different ways. Designs vary widely depending on whether the connection is circuit switched or packet switched. The important fact is that a communications mobility protocol must always be able to get traffic to and from the Mobile Node.