OFDM System for High Data Rate and High Mobility

Download Full-Text PDF Cite this Publication

Text Only Version

OFDM System for High Data Rate and High Mobility

Umesha G B Dr. M N Shanmukha Swamy

Research Scholar Professor & HOD

Department of ECE, SJCE Department of ECE, SJCE Mysore, Karnataka Mysore, Karnataka.

Abstract—In recent years, there has been considerable growth in interest in wireless access to wide area data networks at high data rates. This growth has been fuelled by the simultaneous development of the high-speed wired data networks (e.g., the Internet), wide-spread availability of powerful mobile computing devices, rapidly dropping prices for increasingly sophisticated electronic devices, plus the global deployment of wireless voice systems. High-mobility wireless data access at near-LAN rates is likely to be an attractive option. The harsh cellular radio environment creates significant challenges when trying to provide access at high data rates, especially if high-mobility and wide-area availability are expected even more so when considering the constraints placed on the system design by terminal size and power constraints. This research work focuses into Orthogonal Frequency Division Multiplexing (OFDM) techniques addressing high data rate access in representative mobile cellular-like environments.

General Terms:- Quadrature amplitude modulation, FFT Size, Guard interval, Pilot Subcarriers.

Keywords OFDM, cyclic prefix, FFT, IFFT, and Spectrum.


    A few years ago, high-speed access to the few specialized wired data networks that existed was exclusively available to large organizations. Today, small businesses and end users routinely access the Internet at 1 Mb/s and faster through DSL and cable modem networks [1].The computing power of today's palm-size PDA or a notebook- size computer recently required a room full of hardware [2, 3]. Mobile voice communications taken for granted in today's cellular telephones was, until recently, too bulky and expensive for all but a small handful of users. All of these trends will continue and have fuelled the interest in access to wide area data networks at high data rates with the convenience of high-mobility wireless access [4,5].While high-mobility, fast wireless network access is very attractive, designers must address the harshness of the radio environment. Available spectrum is limited. Multipath dispersion, fading, power restrictions, andsystem complexity create further challenges, especially for small, battery operated, and portable terminals.Clientserver communications dominates computing applications today. The bulk of the traffic flows from the server to the client, with a few user keystrokes or mouse clicks generating large amounts of multimedia server traffic [7, 8]. Even for users creating substantial uplink traffic, there are generally more traffic consumers than generators. In aggregate, downlink flows dominate. Terminal implementation issues provide a

    second reason to study the downlink. Generally, the portable or handheld devices envisioned (e.g., PDAs, notebook PCs or handsets), will be battery powered. Increasing uplink data rates by orders of magnitude requires increases in portable transmit power to meet the link budget, a major challenge with foreseeable battery technology [13, 14]. However, readily applicable downlink techniques can attain the high peak data rates desired.


    Four major impairments to the operation of wireless communications networks are, delay spread, noise, interference, and channel variation. Noise and interference are managed by appropriate budgeting of link margin and frequency reuse, but delay spread is inherent to the outdoor multipath environment. Channel variation is likewise inherent to outdoor mobile operation, with the speed of variation directly related to vehicle velocity and operating frequency. Two system parameters dictate achievable data rates, symbol rate and transmit constellation complexity. Unfortunately, both are constrained by the wireless Channel. The key elements of OFDM systems are described below.A complex constellation with a large number of potential transmitted signals makes the signal more susceptible to channel noise and interference. Efficiently implementing a complex constellation requires amplitude variation, increasing susceptibility to path loss and rapid fading. The channel similarly limits symbol rate. Scarce RF spectrum restricts the emitted bandwidth. The dispersive multipath channel also limits the signaling rate. With the multipath delay in a typical urban environment, delayed copies of a transmit pulse arrive microseconds after the original signal. Severe Intersymbol interference (ISI) occurs when thetransmit symbol interval is not substantially longer than the delay spread. ISI is countered by channel equalization, but rapid variation of the mobile channel complicates this. High symbol rates require long equalizers with high-speed signal processing, even for known channels. In contrast to single carrier systems, OFDM counters the high-mobility outdoors channel by transmitting a simple constellation at low symbol rates. In this way, the OFDM signal is inherently resistant to noise, interference, and delay spread.

    Figure1: OFDM Spectrum.

    Low symbol rates and simple constellations would appear to restrict attainable data rates for OFDM. By transmitting a large number of separate carriers, each conveying a low data rate stream, OFDM can attain high data rates. By properly choosing the symbol duration and carrier spacing, efficient modulation and demodulation of the signal is possible. In particular, using the Fast Fourier Transform (FFT), efficient transmitter and receiver structures are easily attained. Figure 1 illustrates a representative spectrum for an OFDM signal. The carrier spacing, f, is small compared to the frequency selective fading effects so it appears that each carrier experiences flat fading.

    Figure 2: OFDM Transmitter.

    Figure 2 illustrates a block diagram for a typical OFDM transmitter. User data is coded, interleaved, and combined with appropriate control and signalingsignals. The resulting bit stream modulates a series of carriers. Here, modulation simply maps groups of bits to a series of constellation points, represented as complex numbers input to the Inverse FFT. The IFFT transforms the modulated carriers into a sequence oftime domain samples. The IFFT size is generally a power of two for efficiency. Since the number of modulated carriers must be smaller than the size of the FFT to make the analog reconstruction filters realizable, the higher frequency carriers are set to zero.If the time samples were transmitted at this point, the OFDM would not be resistant to delay dispersion and ISI. Delayed versions of previous symbols would mix with the current symbol to degrade the demodulated carriers. The OFDM transmitter adds a cyclic extension to the transmitted samples to counter this. Successive samples are copied from the beginning (and/or end) of the transformed samples to the end (and/or beginning) of the block. Any set of N (N is the FFT size) successive samples drawn from the block can be transformed back to the original modulated carriers with only a phase shift, due to the time

    shifting property of the FFT. Generally, the overall length of the cyclic extension is chosen to exceed the maximum expected delay spread on the channel. Additional time domain processing performed by the OFDM transmitter includes windowing and peak control.

    Windowing is necessary to constrain the transmitted signal to the minimum necessary bandwidth needed for the set of transmitted carriers. By controlling the rise and fall tims of the transmitted signal, little more than the bandwidth needed for the collection of transmitted carriers is needed. The OFDM waveform is equivalent to the summation of a large number of sinusoids of arbitrary amplitude and phase. As such, it may theoretically attain high peak levels, relative to the RMS value. While very high peak levels are rarely attained, peak control is needed to minimize linear power requirements of the transmit amplifiers. Peak clipping, followed by waveform filtering often bhe structure of the OFDM transmitter directly. IQ baseband analog signals or digital IF signals are input to the receiver. A complex baseband rotator removes carrier offset caused by transmit and receive local oscillators differences. After selecting an appropriate sequence of N samples, the time domain waveform is converted to the frequency domain by the FFT. With the correct timing instant, the individual carriers are demodulated and the bit stream deinterleaved and decoded. This OFDM receiver structure performs quiteefficiently when compared to equalized single carrier systems and other techniques.

    Figure3: OFDM Receiver.

    Coding across the OFDM carriers creates a signal that is very robust against the frequency selective fading that might impair or prevent reception of any single carrier. In fact, within the limits of the cyclic extension, coding in frequency complements delay dispersion to create frequency diversity, improving the performance of an OFDM system.By time interleaving across several blocks, short bursts of noise, interference, or rapid fading are also mitigated.


    Two existing standards serve as excellent examples of practical OFDM systems addressing somewhat different environments.

    Table1:802.11a Wireless LAN Parameters.

    User Data Rate




    Data Subcarriers


    Pilot Subcarriers


    FFT Size


    Symbol Duration


    Guard Interval


    Subcarrier spacing

    312.5 KHz

    3 dB Bandwidth

    16.56 MHz

    Channel Spacing

    20 MHz

    Carrier frequency

    -5 GHz

    IEEE 802.11adefines a protocol for indoor high-speed wireless LANs.The European terrestrial digital video broadcast (DVB-T) standard defines an OFDM link for the wide-area environment and addresses high-mobility applications. Table1 lists some important parameters of 802.11a. First, by using 48 data-bearing subcarriers, very high data rates are achieved. Since 802.11a is intended for indoors or, at most, local campus-like operation, the guard time is only 800 ns. If the RMS delay spread is roughly half the guard interval, 802.11a should resist about 400 ns delay spread. This equates to a 400 foot path length difference, representative of an indoor environment. Outdoor reflections from buildings, vehicles, terrain, etc., would create much greater path length differences.Varying distances between 802.11a stations create significant channel quality differences. 802.11a supports link adaptation to attain maximum throughput. A 10:1 range in link throughput is possible with rate1/2 coding and BPSK on the worst channel or rate-3/4 coding and 64QAM on the best. Size, power and cost constraints of a PCMCIA card implementation mean that small, low-cost oscillators must determine the 802.11a modem's carrier frequency. Carrier frequency accuracy becomes an important design issue, especially at 5 GHz where 802.11a wireless LANs operate. Fortunately, a 312.5 kHz carrier spacing relaxes oscillator precision to 20 ppm.Table2 shows the corresponding parameters for the of the DVB-T system. Macrodiversity in Single Frequency Networks (multiple transmitters sending the same information on the same frequency) fills in coverage areas, but creates large artificial delay spread. Thus, DVB-T must support up to 56 µsec

    User Data Rate




    Data Subcarriers


    Pilot Subcarriers


    FFT Size


    Symbol Duration

    231-280 micro sec

    Guard Interval

    7-56 micro sec

    Subcarrier spacing

    4.464 KHz

    3 dB Bandwidth

    7.61 MHz

    Channel Spacing

    8 MHz

    Carrier frequency


    Table2: A subset of DVB-T Parameters

    guard times. Longer guard times require longer blocks to maintain reasonable link efficiencies. Since the active interval of the block is the inverse of the carrier spacing, this translates directly into the smaller 4.464 kHz carrier spacing. With a longer symbol duration, DVB-T needs many more carriers to attain data rates comparable to 802.11a. While 802.11a is intended for burst packet communications, DVB-T uses a continuous downlink, so multi-block coding and interleaving can improve system error rates. With more carriers and longer frames, a complex matrix of pilot carriers is used to allow accurate channel estimation.


    The design of an OFDM system assumes essentially no frequency selective fading within a carrier, so equalization is not needed. In an outdoor environment, where several microseconds of delay spread are likely, this assumption would obviously be violated in802.11a, with the 312.5 kHz carrier spacing. Secondly, it is assumed that the channel is relatively stable within a symbol. For high vehicle velocities, several hundred microsecond long symbols violate this assumption Doppler rates approach several hundred hertz. Our research shows that there is a sweet spot for carrier spacing of several kilohertz, we have focused on this range for wide-area high-mobility applications.

    Table3: OFDM Parameters for 4G

    User Data Rate

    2.56-8.96 Mbps



    Data Subcarriers

    512(Average)(128,256 optional)

    Pilot Subcarriers

    128(Average) (32,64,96 optional)

    FFT Size


    Symbol Duration

    200 micro sec

    Guard Interval

    40 micro sec

    Subcarrier spacing

    6.25 KHz

    3 dB Bandwidth

    4 MHz

    Channel Spacing

    5 MHz

    Carrier frequency

    -2 MHz

    Available spectrum is always a concern for wireless system deployment. OFDM, as other techniques, benefits from frequency diversity inherent in wideband operation yet, often only 5 MHz or less is available. Here, the multicarrier nature of OFDM lends itself well to flexible bandwidth assignment have shown good performance in less than 1 MHz This, plus the portable power burden of wider band operation drive the assumed 5 MHz channelization. Table3 lists a set of representative parameters for an OFDM system we have defined for 4th generation wireless networks. Research convinces us that 2.5 to 5 Mbps is readily attainable, with peak rates of almost 10 Mbps for a sizable fraction of high-mobility users.

    Figure4: OFDM Frame

    As described earlier, time interleaving and frequency coding work together to maximize link reliability. Figure 4 illustrates the time and frequencyrelationship of an OFDM frame. One to four 1 MHz bands are used for OFDM carriers, organized as several OFDM symbols in a 1 millisecond burst, containing data and pilot carriers, coded in time and frequency. With a packet mode downlink, time slots can be assigned to multiple base stations as traffic dictates. This TDMAmode balances the typical user's need for low average data rates against the value of a high peak data rate. It also offers the opportunity for power savings at the mobile, only requiring receiver processing for control messages and traffic for the specific terminal.


    This Research work has described considerations for an OFDM modem intended for asymmetric high data rate wireless network access in wide-area, high-mobility applications.Simulations and real-time prototypes have demonstrated the viability of this approach. Other research work has shown the potential to extend these results to much higher data rates with multiple-input multiple-output (MIMO) smart antenna techniques. Peak data rates of up to 40 Mbps in the wide-area, highmobility environment appear quite feasible with four base antennas and four mobile antennas.


Authors are thankful to Architect Sahana Umesh and Daarya S Umesh for their encouragement and valuable guidance for all the time.


  1. T. Whitworth, M. Ghogho, and D. McLernon, Optimal training and basis expansion model parameters for doubly-selective channel estimation, IEEE Trans. Wireless Commun., vol. 8, no. 3, pp. 1490-1498, 2016.

  2. Z. Wang and G. B. Giannakis, Complex-field coding for OFDM over fading wireless channels, IEEE Trans. Inform. Theory, vol. 49, no. 3, pp. 707-720, Mar. 2003.

  3. Y. Li, H. Minn, N. Al-Dhahir, and A. R. Calderbank, Pilot designs for consistent frequency-offset estimation in OFDM systems, IEEE Trans. Commun., vol. 55, no. 5, pp. 864-877, May 2007.

  4. T. Hwang, C. Yang, G. Wu, S. Li, and G. Y. Li, OFDM and its wireless applications: A survey, IEEE Trans. Veh. Technol., vol. 58, no. 4, pp. 1673-1694, May 2009.

  5. S. He and J. K. Tugnait, On doubly slective channels estimation using superimposed training and discrete prolateshperoidal sequences, IEEE Trans. Signal Processing, vol. 56, no. 7, pp. 3214-3228, July 2008.

  6. T. Y. Al-Naouri, K. M. Z. Islam, N. Al-Dhahir, and S. Lu, A model reduction approach for OFDM channel estimation under high mobility conditions, IEEE Trans. Signal Processing, vol. 58, no. 4, pp. 2181-2193, Apr. 2010.

  7. T. Hwang, C. Yang, G. Wu, S. Li, and G. Y. Li, OFDM and its wireless applications: A survey, IEEE Trans. Veh. Technol., vol. 58, no. 4, pp. 1673-1694, 2009.

  8. Y. Shen and E. Martinez, Channel estimation in OFDM systems, Freescale Semiconductor, pp. 1-15, 2006.

  9. M. K. Ozdemir and H. Arslan, Channel estimation for wireless OFDM systems, in IEEE Commun. Surveys Tutorials, vol. 9, no. 2, 2nd Quarter 2007, pp. 18-48.

  10. Y. Mosto and D. C. Cox, ICI mitigation for pilot-aided OFDM mobile systems, IEEE Trans. Wireless Commun., vol. 4, no. 2, pp. 765-774, 2005.

  11. X. Cai, S. Zhou, and G. B. Giannakis, Group-orthogonal multicarrier CDMA, IEEE Trans. Commun., vol. 52, no. 1, pp. 90-99, Jan. 2004.

  12. Moon, JAeKyoung and Choi, Song In., Performance of channel estimation methods for OFDM systems in multipath fading channels. IEEE Transaction on Communication Electronics. vol. 46, (February 2000): pp. 161-170.

  13. M. Ghogho, D. McLernon, E. Alameda-Hernandez, and A. Swami, Channel estimation and symbol detection for block transmission using data-dependent superimposed training, IEEE Signal Processing Lett., vol. 12, no. 3, pp. 226-229, Mar. 2005.

  14. B.McNair,L.J.Cimini, N.R.Sollenberger, Performance of an Experimental 384 kb/s 1900 MHz OFDM Radio Link In a Wide- Area High-Mobility Environment, Proc. VTC2000, Boston, MA, September, 2000.

  15. The IEEE 802.11 Handbook – A Designer's Companion, IEEE Press, 1999.

  16. J.Chuang, L.J.Cimini, G.Li, L.Lin, B.McNair, N.R.Sollenberger, M.Suzuki, H.Zhao, High-speed wireless data access based on combining EDGE with wideband OFDM, IEEE Comm. Magazine, November 1999.

Leave a Reply

Your email address will not be published. Required fields are marked *