What is OFDMA?

OFDMA or Orthogonal Frequency Division Multiple Access, is a multi-user version of the popular OFDM multi-carrier digital modulation scheme. OFDMA is a modulation scheme that is used by the Wi-Fi 6 (or IEEE 802.11ax) standard. It splits the available channel into smaller sub-bands or sub-carrier frequencies. These subbands are referred to as resource units (RUs) and each resource unit is assigned to an individual user of client, thereby allowing access points (APs) to utilize the RUs to simultaneously serve multiple users.

Benefits in OFDMA Compared to OFDM and FDMA

Improved Bandwidth Efficiency by Supporting Multiple Users at a time

In the traditional OFDM scheme that is used by earlier Wi-Fi standards such as the Wi-Fi 5, each signal or data frame from a user is transmitted sequentially and thus, other users have to wait until the current user has completed transmitting all of the OFDM symbols. While this method contributes to relatively improved spectral efficiency compared to a single user, there exists a latency or delay due to this waiting or contention period. In cases where there are multiple users, the contention period will proportionately increase, resulting in an increased delay.

The OFDMA technique enables APs (Access Points) to communicate with multiple users by optimally assigning them to specific RUs depending on the bandwidth needed, the size of the data, and the channel condition. By dividing the channel and assigning these RUs to multiple users, multiple data frames can be simultaneously transmitted by an AP. Here, each RU is a narrowband channel and hence, the overheads that are generally added to the data frame are also reduced as opposed to OFDM. Also, the contention period is reduced as multiple users are being communicated by a single AP.

Similar to OFDM, the available channel is divided into multiple sub-carriers with each sub-carrier being orthogonal to each other. The word orthogonal means that the null of other sub-carriers will coincide with the peak of any sub-carrier. An illustration is shown in the figure given below.

As a result, multiple users can simultaneously transmit data frames and the AP can still correctly decode all of them. Thus, there is no cross-talk between different users transmitting in different sub-carrier frequencies.

The number of sub-bands or resource units to be used for communication depends on several factors such as user/AP device constraints, data frame size, channel condition, and quality of service (QoS) requirements. For example, in Wi-Fi, the channel bandwidth is 20 MHz. In Wi-Fi 5, a 20 MHz channel is divided into 64 sub-bands with each having a bandwidth of 312.5 kHz. Wi-Fi 6 on the other hand, allows this spacing to reduce to 78.125 kHz, which increases the number of sub-bands to 256. This in turn allows even more users to transmit data frames to a single AP. Thus, OFDMA is designed to utilize the available bandwidth more effectively compared to OFDM and maintains orthogonality between different sub-carriers while allowing multiple users to communicate over these narrow-band channels.

FDMA separates multiple users by allowing a finite amount of spacing in the frequency domain, which inevitably inserts a portion of the spectrum that remains unused, resulting in inefficient use of spectrum. OFDMA orthogonally assigns the RUs to multiple users, thereby utilizing the spectrum more efficiently than the FDMA scheme.

The number of sub-bands that can be used over the available bandwidth also determines the number of users that can connect at a time. As the number of sub-bands increases for a given bandwidth, the sub-band spacing decreases. This number cannot be increased indefinitely since beyond a limit, the sub-band spacing will be too small and result in interference. This relationship puts a limit on the number of users that can operate at a given bandwidth at a time. The table below explains the number of sub-carriers that are available over a given bandwidth and the number of users that can operate at a time.

Resource Unit (RU) Allocation for different number of users, Image Credit: Aerohive Networks       

Multipath Fading and Intersymbol Interference (ISI)

In a traditional single-carrier modulation scheme, a single data frame is transmitted over the entire bandwidth. As we go up in frequency, the bandwidth of the signal also increases. In situations when the signal bandwidth is greater than the channel bandwidth (signal time is less than the channel time), multiple copies of the symbol will interfere with the next consecutive symbol. This interference is due to the multipath reflections – a condition where the signal is reflected off from multiple obstacles in the channel. This interference is called inter-symbol interference (ISI). The below figure illustrates this point.

Effect of ISI due to multipath fading in a channel

In the above figure, the delay spread, or delay time is the channel time and is defined as the time difference between the first and last reflection of the symbol. When this delay spread is less than the symbol time, it will result in ISI and hence, the receiver will find it more difficult to decode the symbols correctly. As a result, the bit error rate (BER) and probability of error increases. Depending on the number of symbols affected and the length of the symbol, the effect of interference and the chances of decoding the symbols correctly will vary. ISI is an undesirable effect that leads to distortion of the original symbol or signal.

Just like OFDM divides the available bandwidth into multiple sub-carriers, each having a narrow bandwidth, OFDMA also divides the channel bandwidth into multiple sub-carriers and where each sub-band or sub-carrier is assigned to a user or client. Therefore, the symbol bandwidth is reduced compared to the channel bandwidth. OFDMA based systems are carefully designed by taking the delay spread that can be expected for any given channel. This means that the symbol time will be greater than the delay spread. This means that the symbol time will be greater than the delay spread and hence, reduces the chances of earlier symbols interfering with consecutive symbols. As a result, ISI is reduced, which in turn reduces the BER. This means that every receiver will be able to decode the data frame correctly even when all of them are simultaneously transmitted and received.

This is not possible with OFDM as multiple users cannot simultaneously transmit over the channel and if transmitted, results in interference due to other users. Therefore, OFDMA offers better resistance to multipath fading and ISI even in the presence of multiple users compared to OFDM. OFDMA also eliminates the use of a guard band, unlike FDMA. In FDMA systems, depending on the severity of the channel, the guard bands have to be increased, thereby resulting in an enhanced multipath fading resistance but at the expense of an increased wastage of bandwidth. OFDMA can utilize the bandwidth more efficiently than FDMA while maintaining a reduced BER.

Furthermore, like OFDM, OFDMA also inserts a guard time interval ensuring that the symbols do not interfere with other symbols in the data frame and symbols from other users. This is done to further minimize the effects of ISI.

Frequency Selective Fading

In OFDMA, the signal (per user) bandwidth is very narrow compared to the coherence bandwidth. Coherence bandwidth is a term often used to describe the range of frequencies in a channel that experiences nearly the same amount of fading. This phenomenon is referred to as flat fading.

If the signal bandwidth is greater than the coherence bandwidth, not all frequency components will experience the same amount of fading. Thus different frequency components of the signal experience different levels of fading, an effect referred to as frequency selective fading. OFDMA systems are optimally designed such that the effects of frequency selective fading are negligible and thus, multiple users can simultaneously transmit over the channel and the receiver can still decode the desired signal correctly.

Key Limitations of OFDMA

Sensitive to Doppler Shifts

Doppler shifts occur when either the transmitter or the receiver is moving relative to each other. This shift causes offsets in sub-carrier frequencies which in turn, changes the orthogonal nature of the sub-carrier frequencies. The impact becomes more severe when the vehicle (on or within which the OFDMA system is mounted) speed increases, causing a larger increase in the sub-carrier offset. When the orthogonality between sub-carriers changes, the effects of ISI will become dominant, resulting in an increased BER. Therefore, OFDMA is not a feasible technique in high-mobility vehicular applications.

Effect of Co-Channel Interference

Co-channel interference is a phenomenon that causes cross-talk when two radio systems are using the same channel. This effect generally arises due to ineffective radio resource management or system design issues. Dealing with this effect is more complex in an OFDMA as the frequency channel is divided among multiple users. Other techniques such as code division multiple access (CDMA) face relatively less complexity in dealing with co-channel interference effects as they require only the codes to be different or orthogonal to each other. In OFDMA, certain sub-carrier frequencies are more prone to experience higher amounts of fading than others. Therefore, a dynamic channel allocation with fast and advanced coordination between base stations is necessary to overcome this effect.

Complex Resource Allocation

The process of scheduling the sub-carrier frequencies to different users will become more complex when the number of users gradually increases over a given bandwidth. Therefore, a faster coordination approach between the APs or base stations is essential to allocate optimal sub-carrier frequencies and the corresponding transmit power for each user. This requires complex hardware and software processing methods, resulting in increased power consumption and cost.