Modelling a high data-rate 802.16 wireless modem

As part of the ITEA MARTES European research project, THALES needed to model a high data-rate 802.16 wireless modem to predict system behavior and performance though system-level modeling and simulation of partial hardware and software, perform design space exploration early in the lifecycle with embedded system executable specifications, and capitalize on the organization’s co-design know-how at a higher abstract level to reduce development cost.

THALES used CoFluent Studio to model an 802.16 modem using the orthogonal frequency division multiplexing (OFDM) modulation techniques to identify adaptations that improve the robustness of the links and/or the output data rate. The modem contains an existing piece of software originally written in C++. It is modular code and uses three freeware libraries and an internal matrix library.

The wireless local area network (WLAN) is a network standard that allows the creation of local wireless networks using free radio frequencies in the 2.4 GHz and 5 GHz spectrum. With a range of few hundred meters, most WLANs are based on the popular 802.11 standard.

Wireless metropolitan area networks (WMAN) of 10 km range are based on the 802.16 standard which addresses frequencies below 11 GHz (non-directional environment, several antennas on large periphery) and on the 802.16a standard for frequencies up to 10 GHz (directional environment). WMAN could potentially compete with 802.11-based solutions in low mobility.

For the MARTES project, THALES modeled a system composed of a transmitter and a receiver as shown in the figure 2.

The DataSource function generates series of bursts of unpredictable bit matrixes. Each bit matrix is then transformed by the Transmitter function, which simulates the processing of a radio device. The Transmitter outputs a matrix of real, complex values, which represent the signal emitted by the antenna. The real part of the values constitutes the I (inphase) channel, and the complex part the Q (quadrature) channel.

These two channels are applied to an analog modulator. The behavior of the analog modulator is not taken into account in the model. The channel function performs operations to simulate the effects of the air channel on the transmitted data: Gaussian noise, multi-paths, Rayleigh scattering. Then, matrixes of bits are transmitted to the sink function. This function traditionally includes the generated simulation data for post-processing analysis. The sink transfers the received data to the compare function, which also receives a copy of the original data emitted by the DataSource function. This comparison module counts the number of errors and calculates the bit error rate.

A bigger version of this graphic is available.

The transmitter module is structured as indicated on the following figure 3. * Coder: the convolutive coder introduces a redundancy into the data to be able to easily find the data emitted, even in case of perturbations on the radio channel.

• Puncture: the puncture function moderates the redundancy introduced previously by eliminating a certain number of bits.
• Interleave: the interleave function mixes the data according to a very precise algorithm to decrease the number of successive identical symbols, reducing the consumption required for the radio transmission.
• Reshape: this function distributes the bits on the frequency sub-bands which are radio frequencies available for the transmission.
• PilotInsert: the pilots are particular sequences inserted on dedicated sub-bands for improving the detection of the radio signal in reception.
• Modulator: the modulator transforms the bits into symbols: complex values (I and Q channels) processed according to a dynamically chosen pattern called constellation (BPSK, QPSK, 16QAM, 64QAM).
• IFFT: the Inverse Fast Fourier Transform function performs the conversion between the frequency domain and the temporal domain required for the output signal.
• PrefixAppend: this function performs the addition of a cyclic prefix to the obtained data.
• The receiver module presents a partially symmetric structure to the transmitter module as presented in the following figure 4.

A bigger version of this graphic is available.

The Depuncture, Deinterleave, PilotRemove, FFT, PrefixRemove functions perform the inverse operations achieved by their counterpart functions in the Transmitter.

The three following modules are more specific:

• ChannelInvert: this function performs the classical channel inversion by canceling certain effects of the air channel based upon measures of the frequency profile of the channel.
• Demap: this function tries to find in the constellation pattern to which bit corresponds the received symbol.
• Decoder: the decoder takes the final decision concerning the value of the received bits using a Viterbi algorithm.