Cellular IP is defined as an extension to Mobile IP. It works locally in a cellular access network. Cellular IP can work with Mobile IP to support wide-area mobility. Cellular IP optimizes the cellular network for fast handovers.11 It also provides integrated mobility control and location management functions at the wireless access points. As Mobile IP manages macro mobility, Cellular IP manages micro mobility.

Architecture

Cellular IP networks are connected to the Internet via gateway routers. Mobile terminals are identified to the network by using the IP address of the BS (access router) as a COA. Because Cellular IP assumes that Mobile IP manages macro mobility, the HA tunnels the IP packets to the gateway router of the Cellular IP network.

Within the network, packets are routed upon the home address of the mobile terminal. In the reverse direction, packets from mobile terminal are routed to the gateway hop-by-hop. After reaching the gateway router, packets are routed through Mobile IP.

Routing

In a Cellular IP network, the gateway router periodically sends a beacon packet to the BSs in the wireless access network. BSs record the interface through which they last received this beacon and use it to route packets toward the gateway. Furthermore, BSs forward the beacon to mobile terminals.

Each BS maintains a routing cache. Packets that are transmitted by MNs are routed to the gateway using standard hop-by-hop routing. Each node in the Cellular IP network that lies in the path of these packets should use them to create and update routing-cache mappings.

This way, routing-cache chain mappings are created, which can then be used to route the packets addressed to the MN along the reverse path. As long as the MN is regularly sending data packets, nodes along the path between the MN’s actual location and the gateway maintain valid routing entries.

Information in the routing cache, which includes the IP address of the mobile and the interface from which the packets arrive, disappears after a certain time, called route time-out. Every consecutive packet refreshes the routing information stored at the network nodes.

Also, a mobile terminal may prevent a time-out from occurring by sending route-update packets at regular intervals, called route-update time. These are empty data packets. They do not leave the Cellular IP networks (i.e., they are discarded at the gateways).

Location Management

Cellular IP uses two caches at each node in the access network. One is the routing cache (already discussed in preceding text). The other one is a paging cache, which is optionally implemented at the BSs.

Although routing cache is primarily used to keep routing information for the ongoing connections, the paging cache is primarily used for idle users. Cellular IP defines an idle MH as one that has not received data packets for a systemspecific time, called active-state time-out.

MNs that are not regularly transmitting or receiving data (i.e., idle nodes) periodically transmit pagingupdate packets to maintain the paging caches, which may be used to route IP packets (when routing-cache mapping for that node is expired).

Paging-update packets are empty packets addressed to the gateway and are distinguished from a route-update packet by their IP-type parameter. These updates are sent to the base station that offers the best signal quality.

Similar to data and route-update packets, paging-update packets are routed on a hop-by-hop basis to the gateway. So, maintaining the paging caches is accomplished similarly to the routing caches, except for two differences.

First, any packet sent by the mobile updates paging-cache mappings, whereas paging-update packets do not update routing-cache mappings. Second, paging caches have a longer time-out than routing caches. Therefore, idle MHs have mappings in paging caches but not in routing caches.

In addition, active MHs will have mappings in both types of cache. All update packets are discarded by the gateway, to isolate Cellular IP-specific operations from the Internet. After the paging time-out, paging mappings are cleared from the cache (e.g., when the mobile terminal is turned off).

Mappings always exist in the paging cache when the MN is attached to the network. If routing-cache mappings do not exist, incoming packets may be routed by the paging cache. However, paging caches are not necessarily maintained in all nodes.

Handovers

In Cellular IP networks, the MN initiates a handover. MHs listen to beacons transmitted by BSs and initiate handover based on signal strength measurements. 8 To perform a handover, an MN has to tune its radio to the new BS and transmit a route-update packet.

These update packets create routing-cache mappings and thus configure the downlink route from the gateway to the new BS.11 During the handover, the MN redirects its data packets from the old to the new BS. At the handover, for a time equal to the routing-cache time-out, packets addressed to the MN will be delivered to both the old and new BSs.

If the wireless access technology allows listening to two different logical channels simultaneously, then the handover is soft. If the MN can listen to only one BS at a time, then the handover is hard (in this case, performances of the handover will be more dependent on the radio interface).

The routing-cache mappings will be automatically cleared the moment the time-out occurs. Two parameters define the handover performances: handover delay (i.e., latency) and packet loss. Handover delay is decomposed into rendezvous and protocol time.

Rendezvous time refers to the time needed for an MN to attach to a new BS after it leaves the old BS. This time is closely related to wireless link characteristics (i.e., the rate of beacons transmitted by the BSs). Protocol time refers to the time spent to restore the connection once the MH has received a beacon from the new BS.

Usually, rendezvous time is small and we may approximate handover delay to protocol time. The second parameter is packet loss during the handover. Packet losses occur as follows: Packets are routed through the old BS until the arrival of the first packets through the new route.

In a hard handover, some packets may be lost in this time interval. These losses are proportional to the handover loop time, which is defined as the transmission time from the crossover node to the old location of the MN, as well as the transmission time from the new location to the crossover node, which is the gateway in the worst case.

Although IP packets may be lost during handover, Cellular IP has lower handover delay than Mobile IP. This is due to the local management of the handover (i.e., only local network nodes should be notified at the intradomain handover).

There is no need for communication with a HA that may be located far away from the MN’s current network. To reduce packet losses during the handover in a Cellular IP network, a possible solution is semisoft handover.

In this case, the routing-cache mappings are created before the actual handover takes place. So, before the handover to a new BS, the MN sends a semisoft packet to the new BS and immediately returns to listen to the old BS.

The idea with semisoft packets is to establish the new route between the gateway and the new BS before the handover execution. During this time, the MN is still connected to the old BS.

After a time period called semisoft delay (e.g., 100 ms), the MN performs a regular handover. The semisoft approach, however, does not ensure a smooth handover. In reality, the transmit time from 80.

Open Issues in Cellular IP

Cellular IP is a protocol and concept that integrates location management functions and fast handovers, which are usually found in today’s mobile systems, with typical Internet routing and addressing mechanisms. Cellular IP solves micromobility, while Mobile IP handles the macromobility. However, there are several open issues.

First, the handover mechanism assures the local management of intradomain handovers (i.e., micro mobility), but it is not impervious to packet losses. Losses disrupt typical Internet traffic, such as TCP flows. Semisoft handover reduces the losses, but still it does not guarantee zero loss.

Second, Cellular IP does not provide mechanisms for QoS support, which is very important for some applications (e.g., real-time services). The protocol is basically proposed for the best-effort service, which is the dominant type of traffic on the Internet today. To be able to support multiple traffic classes with different QoS demands, we should integrate Cellular IP with some of the QoS mechanisms.

Handover Mechanisms for Cellular Wireless Packet Networks

Besides Cellular IP, there are several other proposed solutions to micro mobility as an extension to Mobile IP. We refer to some of them, such as the multicast-based Mobile IPv4 algorithm and IP micro mobility support using Handover-Aware Wireless Access Internet Infrastructure (HAWAII).

There are other micro mobility proposals such as vertical handoffs in wireless overlay networks, hierarchical FAs, as well as recent Internet drafts such as fast handovers for Mobile IPv6 and low-latency handovers in Mobile IPv4.