Adaptive Modulation and Coding for Large Open Office Indoor Wireless Environments Indrakshi Dey * , Geoffrey G. Messier * and S. Magierowski † * Electrical Engineering, University of Calgary, Calgary, Canada, Email: deyi@ucalgary.ca, gmessier@ucalgary.ca † Electrical Engineering and Computer Science, York University, Ontario, Canada, Email : magiero@cse.yorku.ca Abstract—The performance of rate adaptive M-ary quadra- ture amplitude modulation (M-QAM) and adaptive trellis-coded M-QAM in large open office indoor wireless environments is studied in this paper. An appropriate composite fading/shadowing channel model termed the Joint Fading and Two-path Shadowing (JFTS) model is adopted for the indoor wireless environment. Mathematically tractable expressions for the spectral efficiency and average bit error rate (ABER) of adaptive coded and uncoded M-QAM over the JFTS channel are derived. Analytical results demonstrate the performance of adaptive M-QAM over different JFTS channel configurations, and simulation results corroborate the derived analytical ABER expressions. I. I NTRODUCTION In an indoor wireless environment like an open office or laboratory, there are not enough large obstacles to reflect or refract the main waves contributed by the scattering clusters visited by the mobile user. Mobile WLAN users generally restrict their movements to a small area, visiting at most one or two scattering clusters. An appropriate composite fad- ing/shadowing channel model that characterizes the transition from local small-scale fading to global shadowing statistics in a large office indoor environment was proposed in [1] based on an indoor measurement campaign. The proposed model was coined the Joint Fading and Two-path Shadowing (JFTS) model, which combines the Rician fading model and the two- wave with diffused power (TWDP) shadowing model [2]. The JFTS distribution is a convolution of the Rician fading distribution and the TWDP shadowing model. The Rician distribution can be expressed in terms of circular bivariate Gaussian random variable with potentially nonzero mean, while the TWDP distribution is the sum of two half-Ricians. Hence, the JFTS distribution can only be expressed in terms of bivariate noncentralized chi-squared distribution and therefore cannot approach Gaussian statistics. This is unlike conven- tional fading models such as Rayleigh and Nakagami-m that can approach zero-mean complex Gaussian statistics under dif- ferent propagation conditions. This suggests that performance of different communication systems over the JFTS channel model will be very different than what has been predicted by conventional fading/shadowing models and motivates the need for analytical error rate and spectral efficiency expressions specifically for the JFTS channel. In fading environments, the Shannon bound can be achieved by optimizing the transmission technique according to the state of the channel. This technique is known as adaptive modu- lation where optimization can be achieved through adaptive variation of the transmit power level, symbol transmission rate, modulation constellation size, instantaneous bit error rate (BER), coding rate/scheme, or any combination of these parameters [3]–[5]. In particular, rate adaptation through the variation of constellation size has been found to be most effective in improving link spectral efficiency [6]–[9] and easy to implement in practical systems. In this context, uncoded rate adaptive M -quadrature amplitude modulation (M -QAM) has gained considerable interest [8], [9]. To improve the achievable spectral efficiency in the presence of deep fading, adaptive trellis-coded M -QAM (TCM) has been proposed [10], [11], where the coding rate is varied along with the constellation size according to the communication link condition. In this paper, we derive mathematically tractable expres- sions for achievable spectral efficiency and average BER (ABER) of adaptive M -QAM and adaptive TCM over the JFTS fading/shadowing channel model. The analysis provides insights into the performance of adaptive modulation and adaptive coded modulation in indoor WLAN environments. Numerical results corroborate the derived analytical expres- sions. The effects of different JFTS configurations on the spectral efficiency and error probability performance of these adaptive modulation techniques are examined. Some inter- esting observations are made on the performance of coded and uncoded modulations over the JFTS channel as opposed to other conventional channels (e.g., Rayleigh). Furthermore, the performance expressions derived for a general channel model such as the JFTS fading/shadowing model provide the achievable performance measures over a wide variety of practical channel conditions, without assuming that the propagation environment is complex Gaussian distributed. The outline of this paper is as follows. Sec. II describes the channel model considered in this paper. The performance of adaptive uncoded and TCM-coded M -QAM over the JFTS channel is analyzed in Sec. III. Numerical results and discus- sion are given in Sec. IV. Conclusion is drawn in Sec. V. II. SYSTEM MODEL A. The Channel Model We consider the JFTS fading/shadowing channel model [1] proposed to characterize large open office indoor wireless environments. We denote the instantaneous received channel signal-to-noise ratio (CSNR) as γ and the average received CSNR as γ . The probability density function (PDF) of the in-