OFDM (Orthogonal Frequency Division Multiplexing) is a modulation and multiple access technique that has been explored for over 20 years. Only recently has it been finding its way into commercial communications systems. OFDM, or multitone modulation, is presently used in a number of commercial wired and wireless applications. In order for any mobile system to create a rich user experience, it must provide ubiquitous, fast and user-friendly connectivity. OFDM has several unique properties that make it especially well suited to handle the challenging environmental conditions that mobile wireless data applications must operate in.
OFDM is a modulation and multiple access technique that has been explored for over 20 years. Only recently has it been finding its way into commercial communications systems, as Mooreâ„¢s Law has driven down the cost of the signal processing that is needed to implement OFDM based systems.
OFDM, or multitone modulation, is presently used in a number of commercial wired and wireless applications. On the wired side, it is used for a variant of digital subscriber line (DSL). For wireless, OFDM is the basis for several television and radio broadcast applications, including the European digital broadcast television standard, as well as digital radio in North America. OFDM is also utilized in several fixed wireless systems and wireless local area network products. One system, FLASH-OFDM, has been developed to deliver mobile broadband data service at comparable data rates to wired broadband services, such as DSL and cable modems.
It is important that the overall system design be well matched to the service profiles in order to maximize the performance of the system, as well as balance the ultimate user experience it provides relative to the cost to the operator. OFDM enables the creation of a very flexible system architecture that can be used efficiently for a wide range of services, including both voice and data. In order for any mobile system to create a rich user experience, it must provide ubiquitous, fast and user-friendly connectivity. OFDM has several unique properties that make it especially well suited to handle the challenging environmental conditions that mobile wireless data applications must operate in.
CHARACTERISTICS AND EVOLUTION OF DATA FOR A MOBILE ENVIRONMENT
Cellular and PCS communications systems have historically been designed with voice traffic in mind. The patterns associated with voice communications are well known, having been observed since the invention and widespread use of the telephone. Voice can be characterized as relatively predictable, with each party talking about half the time in an interactive manner. The statistics of call duration and time of day are well understood, allowing traffic engineers to use a standard methodology to estimate the amount of capacity needed in a communications system. The wireline telephone network has been engineered in a hierarchical fashion using large circuit switches to efficiently connect one voice user to another. The physical circuit over which a call is made is held open for the entire duration of a call, hence the term circuit switching. Voice in wireline and mobile settings have similar characteristics. Existing cellular telephone systems have therefore been designed in a similar way, and optimized to efficiently provide voice service.
DATA TRAFFIC DIFFERS FROM CIRCUIT VOICE
First, data traffic is much more unpredictable than voice traffic. Data is characterized as bursty, meaning that there is significant variability in when the traffic arrives, the rate at which it arrives, and the number of bits in the messages. Data services are made up of numerous and varied applications, which typically are used over the public Internet or corporate networks and Local Area Networks (LANs). Generally, there are many users simultaneously sharing network resources in a sporadic manner. The network traffic demand that results is bursty in nature with the mix of applications, message sizes, throughput needs, and latency sensitivity. Efficiencies are realized through statistical multiplexing afforded though packet switching.
Second, data has very different requirements in terms of reliability. Whereas voice is very robust and capable of being understood even in a noisy environment with a high bit error rate well above 1%, data applications require extremely reliable delivery, with virtually no tolerance for bit errors. Because some bit errors are unavoidable on wireless links, it is important that fast and efficient recovery schemes are implemented to get the correct data bits to the application. A combination of forward error correction (FEC) and fast acknowledgements (ARQ) satisfies this need. Powerful FEC is employed to dramatically reduce the bit error rate (BER), and ARQ is used to guarantee reliable delivery.
Third, data traffic encompasses a much different and wider range of services than voice. While voice can be considered a data service via Voice over Internet Protocol (VoIP), data services as a whole will span a much broader and more diverse range. Different types of services have different requirements along several dimensions. A data service can be characterized by its importance or priority. This is determined by the quality of service (QoS) that is required, which can be measured by the amount of delay that a user is willing to tolerate, and the reliability required. A service provider may offer differing service rates, or classes of service, accordingly. Premium service users may be given priority over best-effort users, whose traffic is sent if there is capacity available at the time.
MOBILE, WIRELESS DATA OVERVIEW
Todayâ„¢s data traffic is primarily driven by wireline users and generated by a broad range of sources and applications, including Internet use, electronic messaging, file transfer and voice traffic, such as Voice over IP (VoIP) through desktop and portable computers. The TCP/IP protocol suite, which is the data transmission protocol used for the Internet, is the most widely used protocol, and governs the bulk of all data traffic.
In the future, the number of devices generating data traffic is expected to skyrocket by an order of magnitude. Information appliances and other data centric wireless and consumer products will fuel this growth.
Mobile wireless communications has traditionally posed a difficult performance challenge for TCP/IP protocols. TCP was designed and optimized around reliable wireline links, where bit and packet error rates are substantially lower than that typically achievable wirelessly. When TCP encounters dropped or lost packets, it assumes that there is congestion on the link, rather than the link itself being unreliable. Congestion is handled by reducing the information rate that the sender is allowed to transmit at. By interpreting the unavoidable errors that occur in a wireless environment as congestion, the effective data rate seen by the end user is reduced. This is further compounded by the fact that the initial data rate at the start of a TCP session is low â€œ far below the ultimate peak rate .This slow start aspect of TCP can dramatically add to latencies as the link is throttled down due to errors and then slowly ramped back up. The ultimate throughput of TCP is also dependent on the delay associated with ACK packets. The round trip time (RTT) latency and latency jitter directly impact the obtainable throughput as demonstrated
In summary, there is a much wider range of requirements and characteristics for data communications than there is for voice. This variability prohibits data from being efficiently carried over the hierarchical networks designed for voice traffic, whether wireline or wireless. Mobile data systems face additional challenges as a result of the vagaries of the wireless environment.
Mobile cellular wireless systems operate under harsh and challenging channel conditions. The wireless channel is distinct and much more unpredictable than the wireline channel because of factors such as multipath and shadow fading, Doppler spread, and time dispersion or delay spread. These factors are all related to variability introduced by the mobility of the user and the wide range of environments that may be encountered as a result.
Multipath is a phenomena that occurs as a transmitted signal is reflected by objects in the environment between the base station and a user. These objects can be buildings, trees, hills, or even trucks and cars. The reflected signals arrive at the receiver with random phase offsets, since each reflection generally follows a different path to reach the userâ„¢s receiver. The result is random signal fades as the reflections destructively (and constructively) superimpose on one another, which effectively cancels part of the signal energy for brief periods of time. The degree of cancellation, or fading, will depend on the delay spread of the reflected signals, as embodied by their relative phases, and their relative power. Figure 3.1 shows the arrival of a signal and two multipath components.
Time dispersion represents distortion to the signal and is manifested by the spreading in time of the modulation symbols. This occurs when the channel is bandlimited, or in other words, when the coherence bandwidth of the channel is smaller than the modulation bandwidth. Time dispersion leads to intersymbol interference, or ISI, where the energy from one symbol spills over into another symbol, thereby increasing the BER. It also leads to fading.
In many instances, the fading due to multipath will be frequency selective, randomly affecting only a portion of the overall channel bandwidth at any given time. Frequency selective fading occurs when the channel introduces time dispersion and when the delay spread exceeds the symbol period. When there is no dispersion and the delay spread is less than the symbol period, the fading will be flat, thereby affecting all frequencies in the signal equally. Flat fading can lead to deep fades of more than 30 dB.
Doppler spread describes the random changes in the channel introduced as a result of a userâ„¢s mobility, and relative motion of objects in the channel. Doppler has the effect of shifting, or spreading, the frequency components of a signal. The coherence time of the channel is the inverse of the Doppler spread, and is a measure of the speed at which the channel characteristics change. This in effect, determines the rate at which fading occurs.
When the rate of change of the channel is higher than the modulated symbol rate, fast fading occurs. Slow fading, on the other hand, occurs when the channel changes are slower than the symbol rate.
The statistics describing the fading signal amplitude are frequently characterized as either Rayleigh or Ricean. Rayleigh fading occurs when there is no line of sight (LOS) component present in the received signal. If there is a LOS component present, the fading follows a Ricean distribution. There is frequently no direct LOS path to a mobile because the very nature of mobile communications means those mobiles can be in a building, or behind one or other obstructions. This leads to Rayleigh fading, but also results in a shadow loss as well.
MOBILE DATA SYSTEM DESIGN REQUIREMENTS
Conventional wireless systems, including third generation (3G) technologies, have been designed primarily at the physical layer. To address the unique demands posed by mobile users of high-speed data applications, new air interfaces must be designed and optimized across all the layers of the protocol stack, including MAC and networking layers. A prime example of this kind of optimization is found in FLASH-OFDM technology (from Flarion). As its name suggests, the system is based on OFDM, however FLASH-OFDM is much more than just a physical layer solution. It is a system level technology that exploits the unique physical properties of OFDM, enabling significant higher layer advantages that contribute to very efficient packet data transmission in a cellular network. FLASH-OFDM was developed with a number of design objectives, including:
Â¢ Spectrally efficient, high capacity physical layer
Â¢ ?Packet-switched air interface
Â¢ ?Contention-free, QoS-aware MAC layer
Â¢ ?Support for interactive data application including voice
Â¢ Efficient operation using all existing Internet protocols (TCP/IPÂ¦)
Â¢ ?Full vehicular mobility
Â¢ ?Low cost
PACKET SWITCHED AIR INTERFACE
The telephone network, designed basically for voice, is an example of circuit switched systems. Circuit switched systems exist only at the physical layer which uses the channel resource to create a bit pipe. They are conceptually simple as the bit pipe is a dedicated resource, and there is no control of the pipe required once it is created (some control may be required in setting up or bringing down the pipe). Circuit switched systems, however, are very inefficient for burst data traffic.
Packet switched systems, on the other hand, are very efficient for data traffic but required control layers in addition to the physical layer that creates the bit pipe. The MAC layer is required for the many data users to share the bit pipe. The Link layer is needed to take the error prone pipe and create a reliable link for the network layers to pass packet data flows over. The Internet is the best example of a packet switched network.
Since all conventional cellular wireless systems, including 3G, were fundamentally designed for circuit switched voice, they were designed and optimized primarily at the physical layer.The choice of CDMA as the physical layer multiple access technology was also dictated by voice requirements. FLASH-OFDM, on the other hand, is a packet switched designed for data and is optimized across the physical, MAC, link and network layers. The choice of OFDM as the multiple access technology is based not just on physical layer considerations but also on MAC, link and network layer requirements.
TRADITIONAL MOBILE WIRELESS SYSTEM
All modern mobile wireless systems employ a variety of techniques to combat the challenges of the wireless channel. Some techniques are more effective than others, with the effectiveness depending on both the air interface and system architecture approach taken to satisfy the requirements of the services being offered. As mobile systems evolved from analog to digital, more sophisticated signal processing techniques have been employed to overcome the wireless environment. These techniques include diversity, equalization, channel or error correction coding, spread spectrum, interleaving, and more recently, space time coding.
Diversity has long been used to help mitigate the multipath induced fading that results from usersâ„¢ mobility. The simplest diversity technique, spatial diversity, involves the use of two or more receive antennas at a base station, separated by some distance, say on the order of five to ten wavelengths. The signal from the mobile will generally follow separate paths to each antenna. This relatively low cost approach yields significant performance improvement by taking advantage of the statistical likelihood that the paths are not highly correlated with each other. When one antenna is in a fade, the other one will generally not be.
Spread spectrum systems employ frequency diversity. Here the signal is spread over a much larger bandwidth than is needed for transmission, and is typically greater than the coherence bandwidth of the channel. A wideband signal is more resistant to the effect of frequency selective fading than is a narrowband signal since only a relatively small portion of the overall bandwidth will experience a fade at any given time. There are two forms of spread spectrum, code division multiple access (CDMA), and frequency hopping (FH).
CDMA systems, such as those used in IS-95 and 3G WCDMA also employ time diversity in a RAKE receiver. The multipath signals that are received can be time and phase adjusted so that they can be added together as long as the delay is more than one code symbol or chip time. Mix the baseband information stream with a much higher rate pseudorandom spreading sequence code prior to transmission. This effectively increases the signal bandwidth.
A problem that CDMA systems have is that the code sequences are not truly orthogonal in the presence of multipath delay spread. This results in interference between users within a cell. Called multiple access interference, it ultimately limits the capacity of the cell. In fact, more than 2/3 of the interference in a CDMA sector typically comes from the very users in that sector.
Equalization is a technique used to overcome the effects of ISI resulting from time dispersion in the channel. Implemented at the receiver, the equalizer attempts to correct for the amplitude and phase distortions that occur in the channel. These distortions change with time since the channel response is time varying. The equalizer must therefore adapt to, or track, the changing channel response in order to eliminate the ISI. The equalizer is fed a fixed length training sequence at the start of each transmission, which enables it to characterize the channel at that time. A training sequence may also be sent periodically to maintain the equalizers characterization of the channel.
TDMA systems assign one or more timeslots to a user for transmission. There is typically some guard time included between timeslots allow for time tracking errors at the mobile station. The use of equalizers adds to the complexity and costs of the TDMA systems, since equalization requires significant amounts of signal processing power. TDMA systems also have less inherent immunity against multipath fading than spread spectrum systems because they use a much narrower signal bandwidth. On the plus side, TDMA users within a cell are orthogonal to each other since they transmit at different times. Therefore, there is essentially no intra-cell interference.
OFDM FOR MOBILE COMMUNICATIONS
OFDM represents a different system design approach. It can be thought of as a combination of modulation and multiple access schemes that segments a communications channel in such a way that many users can share it. Whereas TDMA segments according to time, and CDMA segments according to spreading codes, OFDM segments according to frequency. It is a technique that divides the spectrum into a number of equally spaced tones, and carries a portion of a userâ„¢s information on each tone. A tone can be thought of as a frequency, much in the same way that each key on a piano represents a unique frequency. OFDM can be viewed as a form of frequency division multiplexing (FDM). However, OFDM has an important special property that each tone is orthogonal with every other tone. FDM typically requires there to be frequency guard bands between the frequencies so that they do not interfere with each other. OFDM allows the spectrum of each tone to overlap, and since they are orthogonal, they do not interfere with each other. By allowing the tones to overlap, the overall amount of spectrum required is reduced.
OFDM is a modulation technique in that it enables user data to be modulated onto the tones. The information is modulated onto a tone by adjusting the toneâ„¢s phase, amplitude, or both. In the most basic form, a tone may be present or disabled to indicate a one or zero bit of information, however, either Phase Shift Keying (PSK), or Quadrature Amplitude Modulation (QAM) is typically employed. An OFDM system takes a data stream and splits it into N parallel data streams, each at a rate 1/N of the original rate. Each stream is then mapped to a tone at a unique frequency and combined together using the inverse Fast Fourier Transform (IFFT) to yield the time domain waveform to be transmitted.
For example, if a 100-tone system were used, a single data stream with a rate of 1 Mbps would be converted into 100 streams of 10 kbps. By creating slower parallel data streams, the bandwidth of the modulation symbol is effectively decreased by a factor of 100, or equivalently, the duration of the modulation symbol is increased by a factor of 100. Proper selection of system parameters, such as number of tones and tone spacing, can greatly reduce, or even eliminate, ISI since typical multipath delay spread represents a much smaller proportion of the lengthened symbol time. Viewed another way, the coherence bandwidth of the channel can be much smaller since the symbol bandwidth has been reduced. The need for complex multi-tap time domain equalizers can be eliminated as a result.
MULTIPLE ACCESS TECHNIQUE
OFDM can also be considered a multiple access technique since an individual tone or groups of tones can be assigned to different users. Multiple users share a given bandwidth in this manner, yielding a system called orthogonal frequency division multiple access, or OFDM. Each user can be assigned a predetermined number of tones when they have information to send, or alternatively, a user can be assigned a variable number of tones based on the amount of information they have to send. The assignments are controlled by the media access control (MAC) layer, which schedules the resource assignments based on user demand.
OFDM can be combined with frequency hopping to create a spread spectrum system, realizing the benefits of frequency diversity and interference averaging previously described for CDMA. In a frequency hopping spread spectrum system, each userâ„¢s set of tones is changed after each time period (usually corresponding to a modulation symbol). By switching frequencies after each symbol time, the losses due to frequency selective fading are minimized. Although frequency hopping and CDMA are different forms of spread spectrum, they achieve comparable performance in a multipath fading environment and provide similar interference-averaging benefits.
Therefore, OFDM combines the best attributes of TDMA, in that users are orthogonal to one another, and CDMA, as discussed above, while avoiding the limitations of each, including the need for TDMA frequency planning and equalization, and multiple access interference(in the case of CDMA).
THEORY OF OFDM OPERATION
The sinusoidal waveforms making up the tones in OFDM have the very special property of being the only Eigen-functions of a linear channel. This special property prevents adjacent tones in OFDM systems from interfering with one another; in much the same manner the human ear can clearly distinguish between each of the tones created by the adjacent keys of a piano. This property, and the incorporation of a small amount of guard time to each symbol, enables the orthogonality between tones to be preserved in the presence of multipath. This is what enables OFDM to avoid the multiple access interference that is present in CDMA systems.
The frequency domain representation of a number of tones, shown in Figure 5.1, highlights the orthogonal nature of the tones used in the OFDM system. Notice that the peak of each tone corresponds to a zero level, or null, of every other tone. The result of this is that there is no interference between tones. When the receiver samples at the center frequency of each tone, the only energy present is that of the desired signal, plus whatever other noise happens to be in the channel.
To maintain orthogonality between tones, it is necessary to ensure that the symbol time contains one or multiple cycles of each sinusoidal tone waveform. This is normally the case since the system numerology is constructed such that tone frequencies are integer multiples of the symbol period, as is highlighted below, where the tone spacing, ?f, is 1/T. Viewed as sinusoids, Figure 6.2 shows three tones over a single symbol period, where each tone has an integer number of cycles during the symbol.
In absolute terms, to generate a pure sinusoidal tone requires the signal start at time minus infinity. This is important since tones are the only waveform than can ensure orthogonality. Fortunately, the channel response can be treated as finite since multipath components decay over time and the channel is effectively band limited. By adding a guard time, called a cyclic prefix, the channel can be made to behave as if the transmitted waveforms were from time minus infinite, and thus ensure orthogonality, which essentially prevents one sub carrier from interfering with another (called intercarrier interference, or ICI).
The cyclic prefix is actually a copy of the last portion of the data symbol appended to the front of the symbol during the guard interval, as shown in Figures 6.1 and 6.3. Multipath causes tones and delayed replicas of tones to arrive at the receiver with some delay spread. This leads to misalignment between sinusoids, which need to be aligned as in Figure 6.3 in order to be orthogonal. The cyclic prefix allows the tones to be realigned at the receiver, thus regaining orthogonality.
OFDM DESIGN CONSIDERATIONS
A number of design tradeoffs must be considered when developing an OFDM-based system. These decisions will be governed by the way the system is intended to be used, including the degree of mobility, the data rates required, the services to be supported, the number of users to be supported, and the environment the system will be used in. The most fundamental tradeoff is the basic sub carrier, or tone characteristics, which involves selection of the number of tones, the bandwidth of each tone, and the cyclic prefix duration.
The cyclic prefix, which is a system overhead, prefix must be long enough to account for the anticipated multipath delay spread experienced by the system. For a given symbol duration the amount of overhead increases as the cyclic prefix gets longer. Delay spreads encountered in cellular systems are typically less than 10 microseconds. The tone spacing, which is reciprocally related to the symbol duration, is an important parameter for mobile system design since it determines the amount of Doppler spread than can be tolerated by the system. The Doppler is a function of the velocity and relative motion of the mobile, but also the frequency of operation. For example, at PCS frequencies, a velocity of 65 mph will result in Doppler of about 250 Hz. This frequency shift can lead to some loss of orthogonality between tones, resulting in interference and reduced performance.
As discussed earlier, most of the physical layer advantages of OFDM are well understood. Most notably, it creates a robust multiple access technology to deal with the impairments of the wireless channel, such as multi-path fading, delay spread and Doppler shifts. Advanced OFDM-based data systems typically divide the available spectrum into a number of equally spaced tones. For each OFDM symbol duration, information carrying symbols (based on modulation such as QPSK, QAM, etc.) are loaded on each tone.
FLASH-OFDM uses fast hopping across all tones in a pseudorandom predetermined pattern, making it a spread spectrum technology. With fast hopping, a user that is assigned one tone does not transmit on the same tone every symbol, but uses a hopping pattern to jump to a different tone every symbol duration. Different base stations use different hopping patterns and each uses the entire available spectrum (frequency reuse). In a cellular employment this leads to all the advantages of CDMA systems, including frequency diversity and out of cell (intercell) interference averaging --- a spectral efficiency benefit that narrow band systems like conventional TDMA do not have.
Different users within the same cell use different resources tones) and hence do not interfere with each other. This is similar to TDMA where different users in a cell transmit at different time slots and do not interfere with one another. In contrast, CDMA users in a cell do interfere with each other, increasing the total interference in the system. FLASHOFDM therefore has the physical layer benefits of both CDMA and TDMA and is at least three times more efficient than CDMA. In other words, at the physical layer, FLASH-OFDM creates the fattest pipe of all cellular technologies. Even though the 3x advantage at the physical layer is a huge advantage, the most significant advantage of FLASH-OFDM for data is at the MAC and link layers.
MAC AND LINK LAYERS
FLASH-OFDM exploits the granular nature of resources in OFDM to come up with extremely efficient control layers. In OFDM, when designed appropriately, it is possible to send very short amount (as little as one bit) of information from the transmitter to the receiver with virtually no overhead. Hence, a transmitter that is previously not transmitting can start transmitting, as little as one bit of information and stop, without causing any resource overhead. This is unlike CDMA or TDMA, where the granularity is much coarser and just to initiate a transmission wastes a significant resource. Hence, in TDMA for example, there is a frame structure and whenever a transmission is initiated, a minimum of one frame (a few hundred bits) of information is transmitted. The frame structure does not cause any significant inefficiency in user data transmission as data traffic typically consists of a large number of bits. However, for transmission of control layer information, the frame structure is extremely inefficient, as the control information typically consists of one or two bits but requires a whole frame. Not having a granular technology can therefore be very detrimental from a MAC and link layer point of view.
FLASH-OFDM takes advantage of the granularity of OFDM in its control layer design enabling the MAC layer to perform efficient packet switching over the air, and at the same time providing all the hooks to handle QoS. It also supports a link layer that uses local (as opposed to end-to-end) feedback to create a very reliable link from an unreliable wireless channel, with very low delays. The network layers traffic therefore experiences small delays and no significant delay jitter. Hence, interactive applications like (packet) voice can be supported. Moreover, Internet protocols like TCP/IP (transport control protocol) run smoothly and efficiently over a FLASH-OFDM airlink. TCP/IP performance on 3G networks is very inefficient because the link layer introduces significant delay jitter so that channel errors are misinterpreted by TCP as network congestion and TCP responds by backing off to the lowest rate.
Packet switching leads to efficient statistical multiplexing of data users and helps the wireless operators support a much higher number of users for a given user experience. This, together with QOS support and a 3x fatter pipe, allows the operators to profitably scale their wireless networks to meet burgeoning data traffic demand in an all-you-can eat pricing environment.
This paper highlights the unique design challenges faced by mobile data systems, resulting from the vagaries of the harsh wireless channel, the wide and varied service profiles that are enabled by data communications, and the performance of wireline based protocols, such as TCP/IP (with the realities of wireless links). OFDM has been shown to address these challenges and be a key enabler of a system design that can provide high performance mobile data communications.
OFDM is well positioned to meet the unique demands of mobile packet data traffic. But in order to seamlessly unwire all the IP applications inherent in the wired Internet and Intranets (including interactive data applications and peer-to-peer applications), all layers of the OFDM air interface need to be jointly designed and optimized from the ground up for the IP data world. This means not to rely solely on OFDMâ„¢s physical layer advantages, but to leverage them into all of the higher layers of the system.