May 2001

BY KENT LEUNG

"Blood type AB, blood pressure's 88 over 58, pulse 130 and dropping," the EMT relays to her partner as the ambulance races down the narrow city streets with its sirens blasting. At a nearby hospital, doctors in the emergency room prepare for the arrival of the patient. Through a wireless Internet connection to the hospital network, the ambulance medics can access the patient's complete health record while keeping doctors in the ER apprised of any changes that could jeopardize his chance of survival. When the doctors receive the patient ten minutes later, they have received updates on his heart rate, breathing, and other vital signs transmitted downloaded from the ambulance Internet in real time.

This scenario is not far from reality, as Mobile IP technologies are enabling devices like patient monitors, voice over IP phones, and computer laptops to communicate seamlessly while traveling between different locations or networks. In an era when maintaining continuous connectivity while roaming has become increasingly important to every day life, it has become even more crucial for mission-critical fields such as the emergency medical profession and the armed services. So, why Mobile IP? Why it is important? How does it work? Furthermore, what are some of the future applications for "networks in motion?"

WHY MOBILE IP?
Mobile IP, an open standard (RFC 2002) specified by the Internet Engineering Task Force (IETF) in 1996, is the technology that enables users to maintain connectivity while roaming between IP networks. IP addressing and routing in networks are fixed at locations, so a device on a network is reachable because it has an address on the network. The IP address is analogous to a phone number associated with a telephone jack. If a phone remains plugged into a particular jack, it is reachable via the phone number assigned to that jack.

The problem occurs when a person having a phone conversation attempts to move to a different location and plug the same phone into a new jack. Because the circuit was designed to function with the original jack, the conversation ceases. IP networks encounter the same challenge. When a device is no longer associated with its former IP address on its home network, its active sessions are dropped. Mobile IP was created to enable users to retain the same IP address while traveling to a different network, thus ensuring that a roaming individual could continue communications without dropping the sessions or connection.

MOBILE IP: THREE-PART HARMONY
Mobile IP is comprised of three entities: The Mobile Node, the Home Agent, and the Foreign Agent. A Mobile Node is a device capable of performing network roaming. Examples of Mobile IP clients are cell phones, PDAs, or laptops whose software enables roaming capabilities. The Home Agent is a router on the home network serving as the anchor point for communications with the Mobile Node; it tunnels packets to the roaming Mobile Node. The Foreign Agent is a router that functions as the Mobile Node's point of attachment when it travels to a foreign network, delivering packets from the Home Agent to the Mobile Node.

Common terminology in Mobile IP also includes the Care-of Address and Correspondent Node. The Care-of Address is the termination point of the tunnel toward the Mobile Node when it is not on its home network, while the Correspondent Node is the device that the Mobile Node is communicating with, such as a Yahoo! Web server.

AGENT DISCOVERY
To facilitate an "always-on" connection for Mobile Nodes roaming between different networks, Mobile IP begins with the process of agent discovery.

The Home Agent and Foreign Agent continuously advertise their services on the network. The Mobile Node, upon receiving the advertisements, discovers these agents and their offered services, and learns whether it is at home or if it has moved to a foreign network. Within the advertisement, the agent specifies whether it is a Home Agent, Foreign Agent, or both; its Care-of Address; the types of services it will provide, such as Reverse Tunneling; and the allowed registration lifetime or roaming period, which may be extended for the Mobile Node.

When the Mobile Node hears a Foreign Agent advertisement, it detects that it has moved outside of its home network and begins registration.

REGISTRATION
To formulate its registration request, the Mobile Node utilizes the information captured within the Foreign Agent's advertisement, protecting the integrity of this data with its Home Agent-shared key before sending the request to the Foreign Agent. As the liaison between the Mobile Node and the Home Agent, the Foreign Agent then processes and relays the request to the Home Agent, which confirms the validity of the Mobile Node using the same shared key for authentication. The Home Agent subsequently constructs a mobility binding, which maps the Mobile Node to the Care-of Address -- the location where the Mobile Node now actually resides within the foreign network.

Mobile IP provides two options for obtaining a Care-of Address: The Mobile Node can procure it from a Foreign Agent or, alternatively it may directly acquire a collocated Care-of Address, which actually represents its current position on the foreign network. A Mobile Node that obtains a collocated Care-of Address will consume an address on the foreign network, while a node that facilitates the process through a Foreign Agent can share the address associated with that agent with other Mobile Nodes.

Because registration automatically expires on both the Home Agent and Foreign Agent, the Mobile Node re-registers to maintain its attachment on the foreign network. The Mobile Node can also de-register to explicitly notify the agents that it is no longer roaming.

In order to transport packets between the Mobile Node and its home network, the Home Agent creates a tunnel to the Care-of Address, and then sends a registration reply back to the Foreign Agent. The Foreign Agent -- or Mobile Node, in the collocated Care-of Address case -- also constructs a tunnel to the Home Agent for successful registrations, and relays the reply to the Mobile Node. Finally, the Mobile Node authenticates this reply, confirming the agent's awareness that it is roaming. Registration sets up the routing for transporting packets to and from the Mobile Node, a process that is accomplished using tunneling.

TUNNELING
The Mobile Node sends packets using its home IP address, effectively maintaining the appearance that it is always on its home network. Thus, even while the Mobile Node is roaming on foreign networks, its movements are transparent to Correspondent Nodes.

Data packets addressed to the Mobile Node are routed to its home network, where the Home Agent now intercepts and tunnels them to the Care-of Address toward the Mobile Node. Tunneling consists of two primary functions: Encapsulation of the data packet to reach the tunnel endpoint, and decapsulation when the packet is delivered at that endpoint.

Typically, the Mobile Node sends packets to the Foreign Agent, which routes them to their final destination, the Correspondent Node. However, this data path is topologically incorrect because it does not reflect the true IP network source for the data -- rather, it reflects the home network of the Mobile Node. Because the packets show the home network as their source inside a foreign network, an access control list on the routers in the network called ingress filtering drops the packets instead of forwarding them. A feature called Reverse Tunneling solves this problem by having the Foreign Agent tunnel packets back to the Home Agent when it receives them from the Mobile Node.

NETWORKS IN MOTION
When deployed by an IP device running the necessary software, Mobile IP enables users to seamlessly roam between networks while they benefit from an "always-on" connection. A Mobile Network takes this concept a step further by allowing a cluster of devices on a network to roam without requiring that each device have Mobile Node software.

The Mobile Router can be defined as a router that supports a Mobile Network. This concept of networks in motion is specified in IETF RFC 2002. The Mobile Router functions similarly to to a Mobile Node; the key difference, however, is the fact that the Mobile Router is able to maintain connectivity for an entire network rather than for just a single Mobile IP client. Agent discovery operation remains the same. During registration, the Home Agent creates an additional tunnel to the Mobile Router and the Mobile Router creates a reciprocal tunnel.

When a packet is sent to a Mobile Network, the Home Agent encapsulates the packet twice and sends it to the Foreign Agent. The Foreign Agent then decapsulates and forwards the encapsulated packet to the Mobile Router, which decapsulates the packet and delivers it to the appropriate node for which it is currently maintaining connectivity. To the rest of the network, the node still appears to be located on the Home Agent; however, it exists physically on the Mobile Network of the Mobile Router.

In order to continue communications while roaming, IP devices on the Mobile Network send packets first to the Mobile Router, which typically forwards the packets to the Foreign Agent for routing. Again, ingress filtering on routers in the network will cause packets to be dropped because the data path is topologically incorrect. To prevent this from occurring, the Mobile Router utilizes Reverse Tunneling as well, encapsulating packets from IP devices and sending them to the Foreign Agent, which encapsulates packets again before routing them toward the Home Agent. The Home Agent decapsulates twice before forwarding the original packets to their destinations.

IP NETWORKS UNPLUGGED
Now recalling our ambulance example, the Home Agent resides within the hospital, while the Foreign Agents are attached to wireless access points deployed throughout the city streets, providing wireless connectivity to the ambulance. Inside the ambulance is the Mobile Router, which is attached to a wireless access interface and maintains connections for several different devices on its Mobile Network.

The ambulance's Mobile Network appears to be at the Home Agent, but the network resides physically at the Mobile Router.

This setup enables doctors in the ER to monitor the patient's vital signs without disruption. To access updated statistics from the ambulance, the computer in the ER sends a request through the Home Agent, which forwards the request to the Foreign Agent. The Foreign Agent then transmits the information to the Mobile Router, which finally communicates with the end device -- the monitor, electrocardiogram, or other equipment being used in real time to record the patient's latest status. As the ambulance races toward the hospital, the Mobile Router attaches to new Foreign Agents and sessions continue.

The Mobile Router is especially useful for mission-critical applications because it maintains "always-on" connectivity for multiple devices that users may want to deploy within the same network. It also ensures that any additional devices added to the network will automatically support roaming capabilities.

As Mobile IP technology advances, it will be utilized in a wide range of applications, enabling continuous connectivity not only to single mobile users, but also in vehicles such as ships, trains, buses, and airplanes. By allowing us to be connected anywhere, at any time we choose, Mobile IP and its enhancements will enable us to harness the power of an enormous resource -- the Internet -- anywhere we happen to travel.

Kent Leung is a senior software engineer in the IOS Technologies Division (ITD) at Cisco Systems. Cisco Systems is the worldwide leader in networking for the Internet. Cisco's networking solutions connect people, computing devices, and computer networks, allowing people to access or transfer information without regard to differences in time, place, or type of computer system.

[ Return To The May 2001 Table Of Contents ]

BY SHAUL SHOHAT

The growth of data communications has been as dramatic as that of Jack's legendary beanstalk. Seemingly overnight, IP has emerged as the transport protocol offering the greatest promise for bringing together next-generation voice, data, and video networks. And though today's IP networks are primarily limited to effectively transporting data, advances in IP networking will soon allow the delivery of voice, as well as data and video.

But until all traffic is delivered by way of IP, service providers must develop networks that will support both legacy and next-generation networks.

Wireless technologies provide alternative service providers (as well as incumbents) cost-effective and more efficient ways to bypass the existing bottlenecks in today's legacy local loop. Before building their next-generation wireless IP networks, service providers should consider the following recommendations:

Increase Capacity Seamlessly
Wireless isn't necessarily an end-to-end solution. It is one segment of a system that allows service providers to connect their core network to the in-building networks. Next-generation wireless IP networks should follow the trends and developments in the networking world and should be transparent to the end customer, and in some part, to the carriers themselves. This can be found in modular product designs that integrate multiple applications over various capacities and frequencies, while complying with all major global standards.

Build Scalable Wireless Networks Using A Variety Of Topologies
Though fast and cost-effective deployment is a top priority in building a wireless network, service providers should also make sure they have the ability to change the wireless topology as demands grow and change in the metropolitan ring.

A wireless network that supports multiple topologies allows service providers to solve network challenges on the spot as building arrangements and city structures change. For example, to penetrate a spread business area, a ring or mesh topology is the best way to connect major buildings and customers. Then, with the growth of nearby densely populated business parks, service providers can add stars of wireless connections from the ring nodes. If properly deployed, all these connections should be as transparent as the services delivered over them.

Support Legacy TDM Connections
Service providers should use wireless to push forward next-generation IP-based networks in order to increase their network and wireless efficiency for the growing IP traffic in metropolitan area networks. But, at the same time, networks should not ignore the large installed base of a legacy local loop. Deploying a wireless network provides an evolutionary path from the current local loop network to the next-generation IP network. Any wireless IP solution that gives 10/100Base-T and Gigabit Ethernet connections to customers carrying pure IP traffic, should take into account the need to support legacy time division multiplex (TDM) connections such as T1s/E1s that carry voice or private lines to large business customers.

The dramatic growth of IP as the transport protocol of choice has opened up both challenges and opportunities for service providers interested in bringing together voice, data, and video networks. To progressively implement their networks across all areas of the metropolitan market -- and rapidly generate revenue as the market emerges -- service providers should seek out a modular product design that allows them to seamlessly integrate multiple networks with multiple applications over multiple capacities and frequencies. Those service providers that choose to integrate legacy systems with next-generation IP networks through intelligent network elements are the ones that will surely be the most successful.

Shaul Shohat is director of Broadband Wireless Access at Ceragon Networks Ltd. Ceragon is a global provider of high-capacity broadband wireless systems for next-generation networks, enabling service providers to deliver high-speed Internet access and integrated data, video, and voice services.

[ Return To The May 2001 Table Of Contents ]

BY DAVID KAHN

As the Internet provides increasing opportunity for the use of streaming audio and video files, the need for high-capacity access "on-ramps" becomes paramount. Optical wireless (OW) links provide high-speed, broadband access with no frequency interference, no licensing requirements, and no expensive cable installation. For those ISP operators who have line-of-sight possibilities to local access hubs, optical wireless is a cost-effective means of essentially creating their own backbone by linking all their POP sites.

Optical Wireless History
The first-generation (pre-1960) optical wireless systems were driven largely by military applications because of the security offered by the close confinement of the beams. Second-generation systems were made possible by the emergence of high-speed, semiconductor light sources including LEDs, lasers, and modulators. Current designs of optical wireless systems are far more sophisticated than prior generations. Systems should include comprehensive management tools to compensate for earlier problems like platform shake and limited product lifetime.

No Licensing Required; No Interference
Spectrum is a limited and increasingly scarce resource with the growth of mobile services and the introduction of relatively high-bandwidth IP-based services (3G) to mobile terminals. As with any commodity in short supply, it has become of great value to those who control it (governments), and of great cost to the carriers who lease it.

There are some bands that are unlicensed, but like the CB bands, these tend to be of utility only to the early entrants and become almost useless with mutual interference when they become popular. The situation is further complicated by the regulatory variations around the world preventing any particular spectrum solution from being universal.

In contrast, OW operates everywhere in a part of the spectrum that does not require a license. As OW systems do not normally interfere with each other, congestion is not an issue.

LEDs And Lasers -- Eye Safety
LEDs and lasers both operate in the infrared portion of the electromagnetic spectrum. The prime regions for atmospheric optical communication are around 850 nm. The only significant safety issue relating to OW systems concerns the risk to the eyes of anyone who may look into the beam. These risks are addressed in the IEC 8025 standard that classifies the hazards of laser and LED optical sources. The prime danger is that light may focus on the eye retina and cause lesions and permanent sight impairment. To be unconditionally safe, terminals must conform to a CLASS 1 designation. This permits viewing at any range over any duration even using optical aids such as binoculars. As the optical path of an OW link is generally not under the exclusive control of the operator, systems must be CLASS 1 to avoid health risks and the possibility of litigation. LED systems are generally Class1 while laser-based systems are generally not.

Fog And Rain
There is a great deal of skepticism regarding the range and availability of optical wireless systems, which are perceived to be adversely affected by weather conditions. Fog is the dominant limiting loss mechanism for OW systems. Thick fog can present attenuation in excess of 200 dB/km, limiting operation to 250 meters or less. Airport visibility statistics can be adapted to determine the probability of the attenuation exceeding a given level over an extended period. In many areas of Europe and North America, OW system availabilities in excess of 98 percent are typical over a range of 1 km. In areas where fog is rare, such as Nevada or Hawaii, availabilities can approach 100 percent. In general, the range of an optical wireless system limited by fog is a little beyond the visibility range.

Conclusion
Optical wireless is a powerful new addition to the arsenal of tools available to the supplier of high-bandwidth connectivity to end-users and to operators of peer-to-peer networks. While it is not a universal solution, requiring visibility over a line-of-sight, it offers a unique blend of fiber-equivalent capacity with robust and reliable performance, low costs, and rapid, flexible deployment.

David Kahn is vice president of product development at Plaintree Systems. Plaintree is a leading manufacturer of optical wireless links, network switches, and LAN and telecommunications products.

[ Return To The May 2001 Table Of Contents ]