Broadband Wireless Access

The IEEE 802.11 Wireless LAN Standards
The IEEE 802.11 wireless LAN standards specify wireless connectivity for fixed, portable, and moving users in a geographically limited area. The switching mode is packet-based. Like the IEEE 802.3 (Ethernet) standard, the IEEE 802.11 standard supports multiple physical layer media.

The 2.4 GHz IEEE 802.11 Standard
The 802.11 physical layer specification allows three transmission options, namely:

• Direct-sequence spread spectrum
• Frequency-hopping spread spectrum
• Diffuse infrared.

Table 1:  Specifications for the 2.4 GHz IEEE 802.11 physical layer

Note that spread spectrum is not used as a multiple access technique in 802.11 wireless LANs. Rather, it is used to protect data signals against the effects of multipath and other propagation impairments.

The standard defines two medium access control (MAC) protocols. The distributed coordination function (DCF) employs CSMA with collision avoidance (CSMA/CA) for contention-based multiple access. Contention-free service is provided by the point co-ordination function (PCF), which is essentially a polling access method. The PCF relies on the asynchronous access service provided by the DCF.

The 5 GHz IEEE 802.11 Standard
The IEEE 802.11 committee has also finalized a 5 GHz wireless LAN standard (approved as IEEE 802.11a on September 16, 1999) that will support wireless data rates of 6 to 54 Mbit/s based on coded orthogonal frequency division multiplexing (OFDM).

Table 2:  Specifications for the 5 GHz IEEE 802.11 physical layer

OFDM has been chosen due to its excellent performance in highly dispersive channels. OFDM also allows considerable flexibility in the choice of different modulation methods. The channel spacing of 20 MHz is a compromise between having high data rates per channel and having a reasonable number of channels in the allocated spectrum.

Out of the 52 subcarriers in each channel, 48 are subcarriers carrying user data while the remaining four subcarriers are pilots that facilitate phase tracking for coherent demodulation. Each subcarrier serves as one communication link between the access point and the mobile terminals. The 800 ns guard interval is sufficient to enable good performance on channels with delay spread of up to 250 ns.

BPSK, QPSK, and 16-QAM are the supported subcarrier modulation schemes with 64-QAM as an option. Forward error correction is performed by convolutional coding with rate of 1/2 and constraint length of 7. The three code rates of 1/2, 2/3, and 3/4 are obtained by code puncturing.

The HiperLAN Wireless LAN Standards
The HiperLAN standards are European standards that support high performance radio-based LANs.

The HiperLAN Type 1 Standard
HiperLAN Type 1 operates between 5.15 and 5.30 GHz at a data rate of 23.5 Mbit/s. The 5 GHz band is compatible with the U-NII band in the U.S. Unlike the 802.11 standard, which supports both the ad-hoc (distributed) and infrastructure (centralized) topologies, HiperLAN Type 1 supports only the ad-hoc topology. However, the standard also caters for the multihop ad-hoc topology, as opposed to the single-hop topology adopted by 802.11. This allows HiperLAN Type 1 networks to be implemented without the need for frequency planning.

HiperLAN Type 1 defines five channel access priority levels according to residual (useful) lifetime and user priority (Table 3). The user priority is an attribute assigned to each packet according to the traffic type it carries. The residual lifetime represents the maximum time interval the packet must be delivered. Since multihop routing is supported, the residual lifetime is normalized to the number of hops it has to travel before reaching the final destination.

Table 3:  HiperLAN Type 1 defines five channel-access priority levels according to residual (useful) lifetime and user priority

The residual lifetime is computed as follows. While a lower priority packet is waiting, its residual lifetime will be decremented. The user may decide to increase the priority of a packet as its residual lifetime decreases. When the residual lifetime becomes zero and the packet has not been serviced, it will be discarded. Within the same priority class, FCFS policy prevails. Hence, the MAC protocol in HiperLAN Type 1 provides either best effort latency for isochronous traffic (e.g., voice, video) or best effort integrity for asynchronous traffic (e.g., data).

The HiperLAN Type 2 Standard
HiperLAN Type 2 allows wireless LANs to be interconnected to virtually any type of fixed network technology. It can carry Ethernet or IP packets, ATM cells, and supports UMTS. The standard provides wireless data rates of up to 54 Mbit/s at the physical layer and up to 25 Mbit/s at the network layer with QoS guarantees such as maximum allowable delay and cell loss ratio. Like the 5 GHz IEEE 802.11 standard, HiperLAN Type 2 is based on OFDM. The physical-layer specifications are similar to those of the 802.11 standard depicted in Table 2. A key feature of the physical layer is the provision of several modulation and coding configurations (Table 4). This allows a HiperLAN Type 2 network to adapt to changing radio link quality.

Table 4:  HiperLAN Type 2 modulation and coding parameters

Unlike 802.11, HiperLAN Type 2 is connection-oriented. Connections must be established between the mobile terminal and the access point prior to data transmission. This is achieved using signaling functions. The connections are time-division multiplexed over the air interface and can be point-to-point and point-to-multipoint. Point-to-point connections are bidirectional whereas point-to-multipoint connections are unidirectional (towards the mobile terminal).

In the HiperLAN Type 2 MAC protocol, the access point exercises centralized control and adapts according to the resources demanded by each mobile terminal. The protocol is based on TDD and dynamic TDMA. TDD allows communication in the downlink and uplink within the same time-slotted frame. The time slots are allocated dynamically depending on the need for bandwidth resources.

The connection-oriented nature of HiperLAN Type 2 allows straightforward implementation of QoS support. Each connection can be assigned a specific QoS in terms of bandwidth, delay, jitter, and bit error rate. It is also possible to employ a simpler approach where each connection is assigned a priority level relative to other connections. The QoS support enables the transmission of a mixture of traffic types of (e.g. voice, video, and data). There are also specific connections for unicast, multicast, and broadcast transmission.

The HomeRF Standard
HomeRF’s Shared Wireless Access Protocol (SWAP) specification defines an over-the-air interface that is designed to support both wireless voice and data traffic. SWAP adopts a hybrid access protocol:

• TDMA for delivery of interactive voice and other isochronous services
• CSMA/CA for delivery of asynchronous high-speed packet data.

Table 5:  Main system parameters for HomeRF

Bluetooth
Bluetooth is a wireless data interface standard that provides a simple means of exchanging data between two portable communications devices (e.g., mobile phones, personal computers). Bluetooth operates in the unlicensed 2.4 GHz ISM band. The standard supports two types of connections:

• Synchronous Connection Oriented (SCO)
• Asynchronous Connectionless (ACL).

SCO packets are transmitted over reserved slots in a point-to-point connection between a controlling master unit and a slave device. Once the connection is established, both master and slave may send SCO packets. A SCO packet allows both voice and data transmission. However, only the data portion is retransmitted when corrupted. The ACL connection supports both symmetric and asymmetric data traffic. The master unit controls the connection bandwidth and decides how much bandwidth is given to each slave. Slaves must be polled before they can transmit data (Figure 1).

Figure 1:  Polling mechanism in Bluetooth

Bluetooth can support one asynchronous data connection, up to three synchronous voice connections, or a connection that simultaneously supports a mixture asynchronous data and synchronous voice. Each synchronous connection supports a data rate of 64 Kbit/s. An asynchronous connection supports 721 Kbit/s in the forward direction while permitting 57.6 Kbit/s in the reverse direction. Alternatively, it can support a symmetric connection of 432.6 Kbit/s.

The efficiency of wireless ATM access varies according to the type of ATM traffic supported. For example, fixed-assignment access is efficient for predictable constant bit rate (CBR) traffic but can result in poor utilization when serving many variable bit rate (VBR) connections. Random access can be inefficient in supporting ATM traffic, due to the unpredictable delays that a lossy wireless channel can induce. On-demand assignment access appears to be a popular choice for wireless ATM since it results in statistical multiplexing that leads to high channel utilization. The drawback is the increased delay and complexity needed to implement a request-reservation mechanism.

MMDS
MMDS operates with a bandwidth of 500 MHz in the 2.150 to 2.682 GHz band and provides large capacities in the order of 10s of Mbit/s (a potential capacity of around 200 video channels).

LMDS
LMDS was originally intended for consumer services with limited interactivity (e.g., digital TV broadcasting, video-on-demand). It was later recognized that LMDS systems have a strong potential to supply broadband services to both homes and businesses and the interest gradually shifted towards these applications.

LMDS typically operates at a millimeter wave band (28 to 31 GHz) and extremely large blocks of allocated spectrum of approximately 1 GHz are available. Thus LMDS promises much larger capacities in the order of 100s of Mbit/s data rates on each link and is capable of supporting emerging broadband telecommunication services including fast Internet access, digital video distribution, video teleconferencing, and other interactive switched multimedia services.

Although channel impairments play a significant role in the design of both MMDS and LMDS systems, LMDS requires special attention since millimeter waves are very much affected by outages due to rain. Thus, the implementation of LMDS demands many innovations in modulation, channel coding, and adaptive antenna techniques.

The customer premise equipment can be attached to LMDS networks using TDMA, FDMA, or CDMA. Currently, TDMA and FDMA are the predominant access methods. These methods apply only to uplink transmissions from the customer site to the hub. The downlink traffic from the hub is based on time-division multiplex.

ITU’s vision of global wireless access in the 21st century, including mobile and fixed access, is aimed at providing direction to the diverse second generation (2G) technologies in the hope of unifying these competing wireless systems into a seamless, 3G radio infrastructure capable of offering a wide range of services. The benefits are enormous since a successful IMT-2000 standard creates a single market for all aspects of cellular telephony.

Coverage Areas and Data Rates
IMT-2000 aims to provide ubiquitous wireless communications in many different environments. It covers both terrestrial and satellite networks, from indoor pico radio cells through outdoor micro and macro cells to satellite mega cells. In addition to international roaming, it will support high data rate services, including 2.048 Mbit/s for indoor users, 384 Kbit/s for pedestrian subscribers, 144 Kbit/s for moving vehicles, and 9.6 Kbit/s for mobile satellite services. The radio technologies are expected to have vastly improved capabilities over existing 2G mobile systems (e.g., multi-environment, multi-mode, multi-band, multimedia operations).

Figure 2:  Convergence of disparate technologies in IMT-2000

CDMA Proposals
A number of Radio Transmission Technologies (RTTs) developed in Asia, Europe, and the U.S. have been proposed for IMT-2000. Three proposals were initially submitted, namely direct-sequence wideband CDMA (W-CDMA), multi-carrier CDMA mode (CDMA2000), and time-duplexed CDMA (TD-CDMA). W-CDMA is backward compatible to GSM and PDC systems in Europe and Asia respectively while CDMA2000 is backward compatible to IS-95 (developed by the TIA) in North America. Recently, the wireless industry has agreed to a harmonized Global 3G (G3G) CDMA framework.

W-CDMA
W-CDMA increases data rates by using multiple 1.25 MHz channels as opposed to the single channel adopted by the current IS-95 standard (Figure 3). A significant concept distinguishing W-CDMA from current IS-95 systems is the introduction of inter-cell asynchronous operation, which is vital for continuous system deployment from outdoors to indoors, and the data channel associated pilot channel for a coherent reverse link as well as a forward link. W-CDMA facilitates the application of interference cancellation and adaptive antenna array techniques on both the reverse and forward links to significantly enhance the link capacity and coverage.

Figure 3:  W-CDMA uplink multirate transmission

Table 6:  W-CDMA parameters

Time-Duplexed CDMA
In TD-CDMA, different channels are multiplexed onto the same time slot. Since the spreading ratio is small, it may require multiuser detection to remove intracell interference. Another reason why multiuser detection is needed can be attributed to the slow power control in TD-CDMA, resulting in highly variable received power levels.

Table 7:  TD-CDMA parameters

CDMA2000
The main goal of CDMA2000 is to provide 144 Kbit/s and 384 Kbit/s with approximately 5 MHz of bandwidth. Like W-CDMA, CDMA2000 employs slotted ALOHA for packet transmission. However, instead of a fixed transmit power, it increases the transmit power after an unsuccessful attempt. This increases the probability of success for a retransmission through the capture effect.

TDMA Proposals
TDMA technology is represented by Universal Wireless Communication (UWC-136), which harmonizes GSM with North America’s TDMA. The CDMA and TDMA air-interface proposals were harmonized through the 3G-Partnership Project (3GPP).

Table 8:  CDMA2000 parameters

A second world standard is expected as GSM and TDMA systems evolve into the 3G standard, Enhanced Data rates for GSM Evolution (EDGE). EDGE increases the data rate of current GSM systems by roughly three times through the use of 8-PSK (3 bits/symbol) modulation instead of GMSK (1 bit/symbol).

Table 9:  Parameters for GSM and EDGE

Many 3G wireless systems involving high-speed wireless LANs, wireless ATM networks, and wireless Internet connectivity are the major focus of recent research efforts. These broadband networks aim to provide integrated, packet-oriented, transmission of text, graphics, voice, image, video, and computer data between individuals as well as in the broadcast mode. Although the underlying access protocols supporting these networks have evolved rapidly, the basic access methods (e.g., ALOHA, CSMA, TDMA, and CDMA) are still very much relevant.