Section 4.4. Methods of Channel Access on Links

4.4. Methods of Channel Access on Links

Voice and video applications that run on wireless networks could require continuous transmission, requiring dedicated allocation of available resources for that application. The sharing of the available bandwidth among applications requiring dedicated resources is called multiple access . But most wireless applications involve transmission of random bursty data, requiring some form of random channel allocation, which does not guarantee resource availability. Multiple-access techniques assign bandwidth to users by allocating a portion of available spectrum to each independent user .

4.4.1. Frequency-Division Multiple Access

In frequency-division multiple access (FDMA), the available bandwidth is divided into nonoverlapping, or orthogonal, slots. In other words, each user is assigned a portion of the channel spectrum. FDMA is the simplest multiple-access mechanism that supports multiple users.

4.4.2. Time-Division Multiple Access

In time-division multiple access (TDMA), each user is assigned a time slot for transmitting. These time slots are nonoverlapping. TDMA techniques are more complex, as they require synchronization in timing among the users. TDMA is also affected by intersymbol interference because of channel distortions, such as multipath delay effects. TDMA divides each channel into orthogonal slots and hence limits the number of users, based on the available bandwidth. TDMA places a hard limit on the number of supported users and bandwidth available for each user. A practical version of TDMA is the random-access technique and is mainly used in local area networks.

Random Access Techniques

Most wireless networks carry traffic in a bursty form. The dedicated resource allocation schemes, such as multiple-access techniques, prove to be inefficient for some types of systems under bursty traffic. With random-access schemes, channels are accessed at random. Aloha-based and reservation-based protocols are two well-known random-access techniques that require packets to be acknowledged . Distortions in the wireless channel of these schemes, however, may result in loss or delay of acknowledgments. In such cases, a packet retransmission is required, which in turn makes the process inefficient. One solution to this problem would be to use smarter link-layer techniques for the acknowledgment packets to increase their reliability. Common objectives of performing channel access are as follows .

• To minimize interference from other users, a transmitter listens before any transmission.

• To give fair access of the spectrum to all other users, a transmitter transmits for only a certain period of time.

• To minimize transmitter power, the same frequency could be reused in farther areas.

In the basic Aloha scheme, each transmitter sends data packets whenever it has data available to send. This naturally leads to a large number of collisions, and hence a number of data packets have to be retransmitted. Hence, the effective throughput of the Aloha channel is very low because the probability of packet collisions is high. The slotted Aloha scheme was developed to deal with the collision problem. In slotted Aloha, the time is divided into slots, and packet transmission is restricted to these time slots. Thus, the number of collisions is reduced significantly. The throughput with slotted Aloha is double that with basic Aloha. Spread-spectrum techniques are used in combination with Aloha to support a large number of users.

Collision detection and carrier sensing are very difficult in a wireless environment. Shadow-fading effects impair collision detection, as objects obstruct the direct signal path between users. This difficulty of detecting collisions in a wireless environment is often referred to as the hidden-terminal problem . Path loss and shadow-fading effects result in signals being hidden between users. Thus, collision-avoidance schemes are normally used in wireless networks, especially in wireless LANs. In a collision-avoidance scheme, the receiver sends a busy tone on receiving a packet. This busy tone is broadcast to all the nearby transmitters. A transmitter that receives a busy tone from any receiver refrains from transmitting. Once the busy tone ends, the transmitter waits for a random amount of time before sending packets. This random back-off scheme is used to prevent all transmitters from transmitting at the same time when the busy signal ends. The collision- avoidance schemes help reduce the collisions in Aloha channels and thus significantly improve the throughput of Aloha.

Reservation-based schemes assign channel to users on demand. The effective channel bandwidth is divided between the data channel and the reservation channel. Users reserve channel bandwidth on the reservation channel and send the data packets on the data channel. Users send small packets along the reservation channel, requesting access to the data channel. If a data channel is available, the request is accepted, and a message is sent to the user. Thus, an overhead in reservation-based schemes is the assignment of the data channel. But these data channels are assigned only on demand. For networks on which only small messages are exchanged, the overhead may be tolerable. Also, when the traffic in the network increases rapidly , the reservation channel may get congested with request messages.

The packet-reservation multiple-access (PRMA) scheme combines the benefits of the Aloha and the reservation-based schemes. In PRMA, time is slotted and organized into frames , with N time slots per frame. A host that has data to transmit competes for an available time slot in each frame. Once a host successfully transmits a packet in a time slot, the slot is reserved for the user in each subsequent frame until the user has no more packets to transmit. When the user stops transmitting, the reservation is revoked , and the user has to compete for a time slot to send more packets. PRMA is used in multimedia applications.

4.4.3. Code-Division Multiple Access

In code-division multiple access (CDMA), users use both time and bandwidth simultaneously , modulated by spreading codes. Spreading codes can be either orthogonal (nonoverlapping) or semiorthogonal (overlapping). Orthogonal spreading codes make it possible to recover the signal at the receiver without any interference from other users. But orthogonal spreading codes place a hard limit on the number of supported users, such as FDMA and TDMA. Semiorthogonal spreading codes involve some interferences from other users at the receiver side. The fundamental superiority of CDMA over other channel-access method lies in the use of the spread-spectrum technique .

Spread-Spectrum Technique

The spread-spectrum technique involves spreading frequencies of a transmitted signal over a wider range. This technique reduces flat fading and intersymbol interference. The message signal is modulated by a pseudonoise signal , which encompasses a wider bandwidth. Hence, the resultant transmission signal obtains a much larger bandwidth.

Spread-spectrum techniques can be implemented in two ways. In the first one, direct sequence , the message signal is Exclusive-ORed, with the pseudonoise sequence thereby spreading the frequency range. The second technique, frequency hopping , involves using the pseudonoise sequence to transmit over a range of frequencies. Frequency hopping is formed based on the random pseudonoise sequence.

Rake Receiver

The rake receiver , shown in Figure 4.4, is used in a CDMA system where multipath effects are common. The binary data to be transmitted is first Exclusive-ORed with the transmitter's chipping code to spread the frequency range of the transmitted signal. The signal is then modulated for transmission over the wireless medium. The multipath effects in the wireless medium result in multiple copies of the signal, each with different delays of t ₁ , t ₂ , and t ₃ and with different attenuations of ± ₁ , ± ₂ , and ± ₃ . The receiver demodulates the combined signal and then introduces a variable delay for each signal. Signals are then combined with different weighing factors ² ₁ , ² ₂ , and ² ₃ to give the resultant signal with reduced multipath effects.

Figure 4.4. The rake receiver at the heart of CDMA

Semiorthogonal CDMA is the most complex type of multiple-access scheme. But in this technique, the mobile units located far from the receiver experience high interference from other users and hence poor performance. Some power-control schemes are used to mitigate this problem by equalization. The semiorthogonal schemes have the advantage that they do not place any hard limit on the number of users supported. But owing to the semiorthogonal codes, the interference from other users increases as the number of users becomes large. However, the interference can be reduced by using steering antennas, interference equalization, and multiuser detection techniques.

In CDMA, the available frequency bandwidth for each cell is divided in half: one for forward transmission between the base station to the mobile unit and the other part for reverse transmission from the mobile unit to the base station. The transmission technique used is called direct-sequence spread spectrum (DSSS). Orthogonal chipping codes are used to increase the data rates and to support multiple users. The transmitted signal also has an increased bandwidth. Using CDMA for cellular systems has several advantages:

• Diversity in frequency. In CDMA, the transmitted signal occupies a wider range of frequencies. Therefore, the transmitted signal is not greatly affected by noise and selective fading.

• Multipath effects. The orthogonal chipping codes used in CDMA have low cross-correlation and autocorrelation. Hence, multipath signals delayed by more than one chip interval do not interfere with the dominant signal.

• Privacy. Since DSSS techniques use pseudorandom sequences, privacy is ensured.

• Scalability. FDMA or TDMA systems support only a fixed number of users. CDMA can support a larger number of users with an acceptable amount of degradation in performance. The error rate increases gradually as the number of users becomes larger.

CDMA also has certain disadvantages. In CDMA, the spreading sequences of different users is not orthogonal, and there is some overlap, which results in some cross-correlation, resulting is self-jamming. Also, signals, being at a greater distance from the receiver, experience significant attenuation compared to signals close to the receiver. Thus, signal strength is weak for remote mobile units.

4.4.4. Space-Division Multiple Access

The three multiple-access techniques FDMA, TDMA, and CDMA are based on isotropical antennas . Recall that an isotropical antenna operates essentially in a uniform manner in all directions. Another method of multiple access in wireless systems is space-division multiple access (SDMA), which uses smart antennas. At one end of communication system, a directional antenna, typically known as a smart antenna, can focus directly on the other end of the system. This technique offers a number of advantages, such as reduction of transmission power, reduced amount of interference owing to reduced transmission power, and the strong signal received by the receiver, owing to the high-gain antenna.

4.4.5. Hybrid Multiple-Access Techniques

An effective multiple-access scheme needs to be carefully chosen on the basis of application requirements. Practical wireless systems normally use two or more multiple-access techniques. This strategy provides a reasonable growth plan and compatibility with existing systems. Multiple-access methods can also be combined to better serve certain applications. The two most commonly used hybrid schemes are FDMA/TDMA and FDMA/CDMA. Other forms of CDMA are so-called W-CDMA and TD-CDMA. W-CDMA provides higher data rates and uses the spectrum in an efficient manner. TD-CDMA combines W-CDMA and TDMA.