SESSION INITIATION PROTOCOL (SIP)

The Session Initiation Protocol (SIP) is an application-layer
control protocol that can establish, modify, and terminate multimedia
sessions or calls. Like MGCP, SIP is text-based. SIP
came out of the Internet Engineering Task Force (IETF) in
1998 as an RFC. It has rapidly gained widespread support,
including Microsoft’s announcement that SIP will be supported
in the next generation Windows XP product.
SIP uses a “request-response” model like that used in
Hypertext Transfer Protocol (HTTP). There is one major difference
between MGCP and SIP—a call agent is not necessary
to mediate between clients. An SIP interface is shown in
Figure 2-19.
The usefulness of SIP for multimedia is almost limitless.
Sessions can be unicast or multicast and include multimedia conferences, distance learning, VoIP, or similar applications.
Some examples of multicast protocols include email, news
groups, Web pages, and the like. SIP also supports the ISDN
and Intelligent Network telephony subscriber services for personal
mobility, which is important for the Wireless Internet.
SIP is reliable, scalable, and can be used with other protocols.
Development is fast because it is very similar to HTTP,
thus making the addition of feature-rich applications very quick
to implement. Initially H.323 and MGCP may be the protocols
of choice for tomorrows’ media gateways and soft switches but
SIP and Megaco will be strong contenders as they mature.
No matter which protocol or protocols become the favorite,
soft switches will be flexible enough to adapt. This adaptability
makes network service providers very happy. Unlike old legacy
switches, this new breed of switches will be quick to accept
added features or changed services without waiting months for
a manufacturer to modify the switch design. 104

MGCP/MEGACO

The Media Gateway Control Protocol (MGCP) specifies communication
between call control elements and telephony gateways.
It is a text-based protocol. Media gateways are telephony
gateways that convert circuit-switched voice signals to data
packets for multiservice packet networks. The Internet
Engineering Task Force (IETF) created MGCP to address
some of the perceived shortcomings of H.323. See Figure 2-18.
The main purpose of MGCP is to place control of call signaling
and processing intelligence in call agents or media gateway
controllers. (Call agents and media gateway controllers are
synonymous with and similar to the gatekeeper functions in
H.323 and are also called soft switches.) A new version of
MGCP, released in August 2000, is called Megaco or H.248.
Although Megaco was created for the same purpose, Voiceover-
IP, it differs from MGCP because it supports a broader
range of networks and devices such as ATM, Remote Access
Servers, Multi-Protocol Label Switching routers (MPLS),
Digital Subscriber Line Access Multiplexers (DSLAMs), and
more.
Because Megaco is very new, interoperability testing is
ongoing. It appears to answer many of the deficiencies of
H.323 and will become very important as we move toward
Voice-over-IP networks. It is somewhat unclear at this time if
Megaco will replace MGCP or just supplement it. Megaco is
more suited for media applications than MGCP, but MGCP
may be a better choice for nonmedia-centric applications, such
as MPLS-based session control.

H.323

H.323 defines packet standards for terminal equipment and
services for multimedia communications over local and wide
area networks communicating with systems connected to
telephony networks such as ISDN. The initial version of this
standard came from the International Telecommunications
Union (ITU) in June 1996.
It defines communication over IP-based local area networks
(LANs). A later version (v2), adopted in January 1998,
extended it over wide area use and general-purpose IP networks.
Several subprotocols are included under H.323 relating
to call setup and signaling.
Four components for a multimedia communication system
as shown in Figure 2-17 include terminals, gateways, gatekeepers,
and multipoint control units (MCU). Gateways and
gatekeepers are used in negotiation for PSTN connections,
whereas MCUs enable multiparty audio and videoconferences.
One drawback of H.323 is that it is somewhat complex and
inflexible. However, it is ISDN-based and relatively easy to
build applications across it. For many applications, H.323 is
satisfactory, but falls short for more advanced implementations
and solutions.
All things considered, the most likely scenario is that multiple
protocols will be used with H.323, such as SIP for
exchange between soft switches and gateways and MGCP for
call setup, because H.323 is too complex and time consuming
to set up a call.

HOME RF

Another industry group, the Home Radio Frequency Working
Group (HRFWG)—made up of members of industry leading
companies such as Compaq, Ericsson, HP, IBM, Intel,
Microsoft, Motorola, and others—created the Home RF
Standard Specification. Home RF combines elements of
802.11 and Digital Enhanced Cordless Telecommunications
(DECT) but supports only up to 2 Mbps. It is aimed at homes
and small businesses.
The price of Home RF is generally less expensive than
802.11 but performance is considerably less. The devices operate
in the 2.4 Ghz ISM band just as 802.11 devices do. In actuality
Home RF competes more with Bluetooth than 802.11. It
was designed for embedded applications in appliances and
computing equipment such as printers. Only time will tell if
this standard prospers.

HIPERLAN2 FEATURES

Other than the high data rate and
QoS features, HiperLAN2 includes including the following:
• Automatic frequency allocation
• Security support
• Mobility support
• Network and application independent
• Power save mode
Automatic frequency allocation is especially important
because this allows for easy installation without the need for
complicated frequency planning such as that required for cellular.
The access points use a built-in support for automatic
transmission frequency allocation.
HiperLAN2 networks also supports authentication and
encryption. A handoff mechanism is managed by the mobile
terminal based on received signals from each access point.
Connections are maintained just in cellular (hopefully maybe
even better). The HiperLAN2 network may also integrated with
a variety of fixed networks.
A power save mechanism is based on mobile terminal-initiated
negotiation of sleep periods. A request is made to the
access point for a low power state and a specific sleep period.
At the end of the sleep period, the mobile terminal searches for
a wake up indicator from the access point, and in the absence
of that, sleeps the next period, etc.

HIPERLAN AND HIPERLAN2

HiperLAN or more recently, HiperLAN2 are standards
approved by the European Telecommunications Standards
Institute (ETSI). HiperLAN2 is the most recent version. It is
an interoperable standard providing high-speed, broadband
connectivity for wireless LANs in corporate environments,
public “hot spots” and home environments
HiperLAN2 provides a 54 Mbps data rate on the globally
allocated 5.15–5.3 GHz band. It also may be used in the
17.1–17.3 GHz band in certain geographic locations. It surpasses
the IEEE 802.11a standard with both greater security
and traffic prioritization capabilities. HiperLAN2 also
includes mechanisms for handoffs between WLANs and 3G
mobile systems.
Currently several European manufacturers are implementing
solutions that provide a wireless Virtual Private Network
(VPN) solution for HiperLAN 2 which includes authentication
and encryption. This will enable wireless mobile users to have
a secure connection to their corporate networks when traveling
through so called “hot spots,” such as airports, hotels and conference
centers.
HiperLAN2 achieves its high data rate by using a frequency
multiplexing method called Orthogonal Frequency Digital
Multiplexing (OFDM) with various physical layer modulation
schemes as shown in Table 2-2.
OFDM is particularly efficient in time-dispersive environments,
i.e. where the radio signals are reflected from many points
such as in offices. The basic idea of OFDM is to transmit broadband,
high data rate information by dividing the data into several
interleaved, parallel bit streams, and let each bit stream
modulate a separate subcarrier. HiperLAN2 is time-division multiplexed
and connection-oriented. It can be used for point-topoint
or point-to-multipoint connections. A dedicated broadcast
channel is also included. Each connection can be assigned either
a simple relative priority level or a specific QoS in terms of bandwidth,
delay, jitter, bit error rate, etc. Hiperlan2 uses an approach
for the Access Channel that differs from the OSI model but is
very similar to the IEEE 802-11 standard as seen in Figure 2-15.

HiperLAN2 was designed for short range communications,
about 150 feet maximum. It is primarily meant to be used in a
stationary environment but does support mobility up to 4.3
feet/second. It may be used on networks with or without infrastructure
to support isochronous traffic such as audio or video
with minimum latency. It can support asynchronous traffic
data of 10Mbps with immediate access. HiperLAN2 is also
compatible with ATM.
Radio-based wireless LANs tend to exhibit randomized
“bursty” traffic patterns which can result in performance
issues. Many factors have to be taken into consideration, when
quality of service is to be measured. Among these are:
• Landscape topography
• Elevations that might cause shadows
• Multi-path from signal-reflection surfaces
• Signal loss through absorbing surfaces
• Quality and placement of the wireless equipment
• Number of stations
• Interference
• Etc.
These and other factors have been figured into the
HiperLAN2 specification to allow for a certain level of Quality
of Service guarantee.
Figure 2-16 depicts a typical topology of a HiperLAN2 network.
The Mobile Terminals (MTs) communicate with one
Access Point (AP) at a time over an air interface. As a user
moves from one AP to the next, handoffs can take place. In an
ad hoc networks, the MTs communicate directly., can also be
created, but their development is still in early phase. The
HIPERLAN/2 is planned to be finalized by the end of 1999.

UNLICENSED SPECTRUM USAGE FOR WLAN

The Federal
Communications Commission (FCC) specifies the rules for
operating in the unlicensed 2.4 GHz spectrum. The largest
governing concern is harmful interference with authorized
services and must work around any interference that may be
received from phones, microwaves or other RF devices.

The FCC mandates that a device must operate in one of
two ways in the 2.4 GHz ISM band:
• Frequency Hopping Spread Spectrum (FHSS). The frequency
changes in a pseudo-random manner based on a predefined
code.
• Direct Sequence Spread Spectrum (DSSS). The data signal
is broken up into sequences and transmitted to the receiver,
which reassembles the sequences into the data signal.
Future versions such as 802.11g may adopt OFDM if the
FCC decides to support it and the industry can agree to rally
behind it. However, at the time of publishing this book, these
are two very big “ifs.”
It is estimated that more than 7.8 million wireless LAN
chipsets were produced in 2000. A similar number is expected
in 2001. Sales are growing from almost $400 million in 2000
to $1.2 billion by 2005. Costs have dropped during 2001, causing
widespread usage in homes and enterprise systems.
However, 2002 will see the release of more Home RF and
802.11g products also. Parks Associates estimates that, while 5
percent of U.S. households currently have a PC network in
place, as many as 15 percent will have one in five years. Of
that, wireless networking will account for 40 percent of all
those home networks.

FEATURES

IEEE802.11 also supports infrastructure networks
and ad hoc networks. One very important characteristic
of 802.11 is that the data rate will be automatically decreased
as signal deteriorates between the access point and the stations.
While 802.11b does include a security mechanism, it has
been discovered to be weak. It also supports station roaming
between access points.

IEEE 802.11

Wireless Ethernet is IEEE 802.11b today, the IEEE standard
for wireless LAN’s. IEEE 802.11b operates in the ISM band at
11 Mbps. However, several new versions of the standard is
being developed, 802.11a, which supports data rates of up to
54 Mbps, and operates in the 5-GHz UNII (Unlicensed
National Information Infrastructure) band. Another version
802.11g is currently being developed which will support up to
20+ Mbps. Table 2-1 summarizes the different versions of
802.11 and includes HiperLAN2 for comparison. It should
also be noted that the IEEE is working on 802.11e, a standard
that spans home and business environments with QoS and
multimedia support while maintaining full backward compatibility
with 802.11b and 802.11a. This version will support
voice and include a higher level of security than 802.11b. The
release date for the standard is unclear at this time.
The IEEE802.11b specification was finalized in 1999 and
quickly adopted by many companies. However, it was just as
quickly discovered that there are two problems: the security is
weak and the theoretical transmission speeds of 11 Mbps falls
short—real world speed is only about 7 Mbps.

BLUETOOTH

Bluetooth is a low-cost, low-power, short-range radio link for
mobile devices and for WAN/LAN access points. It operates in
the ISM band. The Bluetooth standard was created primarily
to replace serial cables between computers and printers or
other peripherals. Speed and reliability were key considerations.
Bluetooth is capable of both voice and data communications
at speeds up to about 70 Kbps.
Bluetooth technology is an enabling technology for the
Wireless Internet and the mobile user. It can be an Internet
bridge between a mobile device and a wireless access point in
an ad-hoc network, as are other WLAN technologies such as
802.11 or Home RF. However, some features of Bluetooth are
unique to it and not available in other WLAN technologies.
Bluetooth actually creates a Personal Area Network. It is small
enough to be embedded in everyday devices such as headsets
or microphones. It can be embedded in a PDA and automatically
synchronize a computer to a PDA. Bluetooth can also
download a file or picture received on a Wireless Internet
phone to a printer or a PDA or computer.
Applications for Bluetooth wireless technology come from
no only the telecom industry but also from the computer, home
entertainment, automotive, health care, automation, and toys
industries. What good is a wireless Internet session if you must
constantly connect to wired network to print? Bluetooth uses a
low-cost short-range radio link or bridge between Bluetooth
enabled devices. Computers, phones, printers, wireless headsets,
and microphones can all communicate with each other
without wires being dragged about. Bluetooth started as an
idea in 1994 at Ericsson. Today, the Bluetooth SIG boasts
almost 2,500 members with nearly every major communications
company represented.
Bluetooth computer and telecom consumer products will
appear in late 2001 or early 2002. Products in other industry
sectors will become available later in 2002.
The Bluetooth Specification addresses two ranges: short
(around 10 m) and medium (around 100 m). The radio link is
capable of voice or data transmission to a maximum capacity of
720 Kbps per channel. The radio spectrum used is in the unlicensed
ISM band at 2.4 GHz. Modulation is Frequency
Hopping Spread Spectrum (FHSS).
Because Bluetooth encompasses many applications, there is
no single competitive technology. Infrared is a competitor in some
cases but it requires line of sight, whereas wireless LANs have
much greater range. Perhaps the closest competitor is Home RF
but it too is more a wireless LAN than a personal area network.