R-3 Repository :: Browsing by Author "Das, Suman"

Browsing by Author "Das, Suman"

Now showing 1 - 19 of 19

Arithmetic Acceleration Techniques for Wireless Communication Receivers
(1999-10-20) Das, Suman; Rajagopal, Sridhar; Sengupta, Chaitali; Cavallaro, Joseph R.; Center for Multimedia Communications (http://cmc.rice.edu/)
We develop techniques to accelerate the implementation of the next generation wireless communication algorithms in hardware. We discuss an implementation of a key computationally intensive baseband algorithm for joint multiuser channel estimation and detection for this purpose and study its real-time requirements. An analysis of the bottlenecks present in the algorithm is made. We present an acceleration technique using task decomposition to take advantage of the existing pipelining and parallelism flow in the algorithm. We show that an application specific system design with multiple processing elements is more effective than the conventional single processor approach as it can satisfy the high data rate requirements of the next generation wireless communication systems. Our analysis is done independent of the final mapping of the processing elements in hardware.
Computationally Efficient Iterative Multiuser Detection and Decoding
(1998-11-20) Das, Suman; Erkip, Elza; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
We describe an iterative detection and decoding scheme for the uplink in a convolutionally coded direct sequence code division multiple access (DS-CDMA) system. We consider maximum-a-posteriori (MAP) decoding and reduce the complexity through an iterative interference cancelation scheme combined with a suboptimal channel decoding algorithm. The MAP priors are updated at every iteration step. We investigate the performance of this low complexity scheme and observe that it is close to optimal.
Computationally Efficient Multiuser Detectors
(1997-09-20) Das, Suman; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
CDMA is becoming an increasingly popular multiplexing scheme in wireless communications and this has necessitated the development of efficient detection techniques. The exponential complexity of the optimal detector on one end and inferior performance of conventional single-user detctor at the other have led to the development of suboptimal multiuser detectors with lower complexity. Most of these detection techniques involve solution of a linear system. In their naive implementation this requires O(nÂ³) operations in the size of the matrix. This cost can be reduced if we move towards modern iterative techniques for solution of the system. However maximum benefit can be achieved if we fully exploit the structure of the system. In this paper we have proposed several methods of reducing the computational complexity utilizing the above ideas. We have also come up with algorighms which computationally can achieve the lower bound in complexity.
Design of Computationally Efficient Multiuser Detectors for CDMA Systems
(1997-05-20) Das, Suman; Center for Multimedia Communications (http://cmc.rice.edu/)
Linear multiuser detectors have been developed for CDMA systems to maintain a balance between the performance and computational complexity of the detectors. However for a system with a large number of users and huge data block length, the cubic complexity of these naive algorithms is computationally prohibitive for real time estimation of transmitted data bits. In this thesis we have proposed methods to reduce the amount of computation by following two approaches. We have divided the problem into smaller subproblems and then solved these subproblems individually. We have found that pipelining and windowing techniques applied to linear feedback detectors can reduce the total amount of computation, without signifficantly sacrificing the performance. Secondly, we have proposed two novel algorithms which exploit the block-Toeplitz structure of the correlation matrix and solve the linear system in only O(NKÂ²) steps. Finally, in a practical system, the delays associated with the users and hence the correlation matrix vary dynamically. We have shown methods to update the solution of the system based on the previous results and the rank of the perturbations of the system. We have also computed a threshold for the number of users whose delays can vary, beyond which fresh computation of the solution of the system is less expensive than the updating scheme.
Design of computationally efficient multiuser detectors for CDMA systems
(1997) Das, Suman; Cavallaro, Joseph R.
Linear multiuser detectors have been developed for CDMA systems to maintain a balance between the performance and computational complexity of the detectors. However for a system with a large number of users and huge data block length, the cubic complexity of these naive algorithms is computationally prohibitive for real time estimation of transmitted data bits. In this thesis we have proposed methods to reduce the amount of computation by following two approaches. We have divided the problem into smaller subproblems and then solved these subproblems individually. We have found that pipelining and windowing techniques applied to linear feedback detectors can reduce the total amount of computation, without significantly sacrificing the performance. Secondly, we have proposed two novel algorithms which exploit the block-Toeplitz structure of the correlation matrix and solve the linear system in only O($NK\sp2)$ steps. Finally, in a practical system, the delays associated with the users and hence the correlation matrix vary dynamically. We have shown methods to update the solution of the system based on the previous results and the rank of the perturbations of the system. We have also computed a threshold for the number of users whose delays can vary, beyond which fresh computation of the solution of the system is less expensive than the updating scheme.
Distorted Channel Capacity: A Geometrical Approach
(2000-06-20) Das, Suman; Cavallaro, Joseph R.; Erkip, Elza; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
Shannon's capacity result has often been used as a benchmark to compare the performance of practical codes. Shannon's theorem helps us to calculate the maximum rate at which data can be transferred without making any errors. Most of the practical codes never achieve this zero error bound. Moreover most of the digital data like audio and video can tolerate various amount of errors. In this paper we introduce the notion of distorted channel capacity - the maximum rate at which data can be transferred when we allow certain amount of error in the recovered bits. We also calculate the distorted channel capacity for a gaussian channel and a fading channel.
Efficient Multiuser Receivers for CDMA Systems
(1999-09-20) Sengupta, Chaitali; Das, Suman; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
We focus on the design of multiuser receiver structures for Code Division Multiple Access (CDMA) communication systems, in the presence of multipath elects and multiple sensors at the base station receiver. We present a exible and extensible framework that allows the use of an estimated elective spreading code from the channel estimation phase, in the multiuser detection process. The effective spreading code captures all the channel parameters such as path delays, attenuation factors, and directions of arrival. Hence estimation of this one composite vector removes the necessity of estimating each individual parameter, thus reducing computational complexity. Our results also show that this approach leads to better performance for multiuser detection, especially when the channel consists of a number of low energy paths in addition to a few discrete strong paths.
Fast Multi-user Detector for a Time-varying CDMA System
(SPIE, 1997-07-01) Das, Suman; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communication
This paper investigates methods to reduce the amount of computation needed to detect information bits using a linear detector for a CDMA system. We show windowing technique coupled with pipelining can reduce the amount of computation without significantly sacrificing the performance of linear feedback detector. We also describe efficient techniques to adapt to a dynamic system where the system parameters vary due to the change in delays associated with individual users.
Fixed Point Error Analysis Of Multiuser Detection and Synchronization Algorithms For CDMA Communication Systems
(1998-04-20) Sengupta, Chaitali; Das, Suman; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
Conventional correlation based single-user techniques for Direct Sequence Code Division Multiple Access (DS-CDMA) wireless communication systems are susceptible to performance degradation due to interference from other users. Recent research has focused on development of several multiuser techniques where information about multiple users is used to improve performance for each individual user. Due to performance benefits of these methods, they are attractive candidates for implementation in future cellular systems. In this paper we present an error analysis of fixed point implementation of some of these techniques.
Hardware design issues for a mobile unit for next generation CDMA systems
(1998-07-20) Das, Suman; Sengupta, Chaitali; Cavallaro, Joseph R.; Center for Multimedia Communications (http://cmc.rice.edu/)
This paper addresses hardware design issues of a mobile receiver for future generation direct sequence CDMA wireless communication systems. In the design of a mobile unit, fixed-point hardware is an attractive alternative because of increased speed, reduced power consumption, and reduced hardware costs. In this paper, we focus on the fixed-point implementation of `blind' detection and channel estimation schemes that do not require knowledge of spreading codes of the other users. The error pattern of the 'blind' algorithms, wordlength requirements, and the operation count required for implementation of such algorithms are evaluated. Our results show that the blind maximum likelihood channel estimation along with the blind MMSE detection algorithm can achieve approximately five times improvement in performance over the conventional correlator based receivers. These newer algorithms require slightly higher wordlength but similar computational complexity.
Iterative Multiuser Detection and Decoding
(1998-11-20) Das, Suman; Erkip, Elza; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
An iterative detection and decoding algorithm is explored for a convolutionally coded DS-CDMA system. The optimum decoding strategy is to consider the trellises of all the users simultaneously, but this has exponential complexity in the number of users. We use an iterative +multistage detection scheme combined with the MAP algorithm to reduce the complexity of the receiver without significant performance degradation. The simulation results show that our algorithm consistently outperforms other existing low complexity algorithms. We also provide a framework where the similarities and differences between various algorithms become transparent. Based on this framework, we make a quantitative comparison of the various schemes under investigation.
Joint Multiuser Channel Estimation and Detection for CDMA Systems
(1999-02-20) Sengupta, Chaitali; Das, Suman; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
We focus on the design of multiuser receiver structures for Code Division Multiple Access (CDMA) communication systems, in the presence of multipath effects and multiple sensors at the base station receiver. The joint channel estimation and detection approach presented in this paper uses a miximum likelihood approach to estimate an effective spreading code [1] vector of each user. We design our channel estimation algorithm to estimate this vector in a computationally efficient manner, instead of focusing on estimation of the individual channel parameters, such as path delays, attenuation factors, and direction of arrival. We present a framework that allows the use of the estimated effective spreading code in the detection process [2] and facilitates the development of a computationally efficient solution to the combined problem of channel estimation and multi-shot detection in a scenario involving multiple users, multiple paths, and multiple sensors.
Low Complexity Iterative Multiuser Detection and Decoding for Real-Time Applications
(IEEE, 2005-07-01) Das, Suman; Erkip, Elza; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communication
This paper presents a low-complexity multiuser decoding technique that can be implemented in real time for a convolutionally coded direct sequence code division multiple access (DS-CDMA) system. The main contribution, denoted here as the iterative prior update (IPU), consists of iterative interference cancellation and prior updates on sequences of coded bits combined with M-algorithm and list decoding. We illustrate performance gains over other low-complexity sequence detection and decoding strategies and argue that the algorithm converges within a few iterations and requires only a small size buffer for keeping track of the priors along iterations. The fact that the we can use existing available architectures for Viterbi decoding with slight modifications and can meet the real-time processing constraints makes the IPU algorithm an attractive alternative for cellular systems.
Maximum Weight Basis Decoding of Convolutional Codes
(2000-06-20) Das, Suman; Erkip, Elza; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
The effectiveness of the convolutional code increases with the constraint length of the code. Unfortunately the decoding complexity of Viterbi algorithm grows exponentially with the constraint length. In this paper we propose a suboptimal decoding method based on the calculation of the maximum weight basis of the convolutional code. We extend the algorithm to incorporate the ideas of list decoding method. The complexity of the algorithm grows only quadratically with the constraint length and the performance of the algorithm is comparable to the optimal Viterbi decoding method.
Maximum Weight Basis Decoding of Convolutional Codes
(IEEE, 2002-11-01) Das, Suman; Erkip, Elza; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communication
In this paper we describe a new suboptimal decoding technique for linear codes based on the calculation of maximum weight basis of the code. The idea is based on estimating the maximum number locations in a codeword which have least probability of estimation error without violating the codeword structure. In this paper we discuss the details of the algorithm for a convolutional code. The error correcting capability of the convolutional code increases with the constraint length of the code. Unfortunately the decoding complexity of Viterbi algorithm grows exponentially with the constraint length. We also augment the maximal weight basis algorithm by incorporating the ideas of list decoding technique. The complexity of the algorithm grows only quadratically with the constraint length and the performance of the algorithm is comparable to the optimal Viterbi decoding method.
Multiuser Information Processing in Wireless Communication
(2000-09-20) Das, Suman; Center for Multimedia Communications (http://cmc.rice.edu/)
Wireless channel is not very conducive towards error-free raw data transmission. On the other hand the tremendous growth in wireless services has made the channel bandwidth a scarce resource and effective utilization of this resource is mandatory. Thus it is instructive to know the limits of a wireless channel. Shannonâ s theorems on channel capacity have been used so far to find the maximum rate at which data can be transmitted over any noisy channel. The theorem calculates the minimum signal to noise ratio (SNR) required to transmit data across a channel with zero probability of sequence error. However the result is practically inhibitive, as it requires encoding and decoding of infinite length code sequences. Practical finite codes never achieve this zero error limit. For practical code design bit-error-rate is often a preferred metric over sequence error rate. However there is no satisfactory method to compare the Shannonâ s capacity results with the bit error rate performance of the practical codes. We introduce the notion of distorted channel capacity to bridge this gap. This measure defines the capacity of a channel when a particular bit-error-rate is allowed. It can also be used as a benchmark to measure the â goodnessâ of a code. Our results show that most of the practical codes lie far beyond the capacity limit. We see that Turbo codes and the convolutional codes come close to this achievable at a prohibitively large computational cost. Specifically, for the convolutional codes the performance improves with large constraint length codes. However the optimal decoding complexity of the convolutional codes grow exponentially with this parameter. We propose a suboptimal decoding technique that has linear complexity in the size of the constraint length and provides close to optimal performance. We further extend our results to a multiuser environment. The optimal joint decoding complexity of multiple users data symbols is exponential in the number of users. Our proposed iterative joint interference cancellation and decoding technique provides computational gain without performance loss.
Multiuser information processing in wireless communication
(2000) Das, Suman; Cavallaro, Joseph R.
The available wireless bandwidth cannot keep pace with the amazing growth rate of wireless subscribers. Moreover, the higher data rate and superior quality of service demand of the increasingly popular multimedia services over wireless networks mandate efficient utilization of the channel. Unfortunately, the wireless channel is not very conducive towards error-free raw data transmission. Forward error correction schemes are therefore necessary for reliable information delivery. Error correcting schemes rely on the addition of redundancy to information content. Therefore it is essential to know the limits of channels in order to design the most efficient wireless systems. Traditional Shannon capacity defines the maximum rate at which data can be perfectly transferred over a channel. But none of the practical error control schemes achieve zero error rate. Moreover, most of the wireless applications can tolerate certain bit error rates. We introduce the notion of distorted channel capacity that measures the maximum rate of data transfer over a channel under a given bit error rate criteria. The performance of the various practical codes benchmarked against this measure reveals that only complex codes can perform close to the optimal bound. For example, convolutional codes of large constraint lengths perform much better than codes with smaller constraint lengths. The decoding complexity of the commonly deployed Viterbi algorithm, however, grows exponentially with the constraint length. We propose a new maximal weight basis decoding technique whose complexity is only quadratic in the constraint length and hence is better suited for real time systems deploying strong codes for error protection. We finally investigate a computationally efficient implementation of a multiuser wireless receiver. The optimal joint decoding complexity of a coded multiuser system is exponential in the number of users. Our proposed iterative joint interference cancellation and decoding technique strikes a balance between the computational complexity and the receiver performance.
Partially Blind Multiuser Detection
(1999-03-20) Das, Suman; Bhashyam, Srikrishna; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
Multiple access interference (MAI) is a significant limiting factor in the performance of direct sequence code division multiple access (DS-CDMA) systems. Various multiuser detection techniques have been developed to combat the effects of MAI. These detection techniques either assume the knowledge of all the users in the system (conventional) or assume the knowledge of the user of interest only (blind). Due to the limitations of the blind algorithms in the presence of a large number of interferers, there is a significant performance gap between these two classes of detectors. Additionally, in practice, the receiver could have only partial knowledge of the interference. In this paper, we develop a new class of detectors, partially blind multiuser detectors, that use information about a subset of interferers and bridge the performance gap between the blind and the conventional multiuser detectors.
Partially Blind Multiuser Receivers for CDMA Systems
(1999-12-20) Das, Suman; Bhashyam, Srikrishna; Cavallaro, Joseph R.; Aazhang, Behnaam; Center for Multimedia Communications (http://cmc.rice.edu/)
Multiple access interference (MAI) significantly limits the performance of code-division multiple access (CDMA) systems. Existing multiuser receivers combat MAI using the knowledge of either all (conventional) or none (blind) of the interfering users. However, in practice, partial knowledge of the interference is often available at the receiver. For example, the base station of a cellular communication system typically has information about the interfering users within its cell only (about 60\\% of the total interference). Therefore, conventional multiuser receivers cannot be used. We show the degradation in performance resulting from treating the unknown (out-of-cell) interference as noise or using the blind receiver. We then develop a new "partially blind receiver" for joint synchronization and detection that combines a nonlinear interference cancellation technique for the in-cell interferers with the linear blind minimum mean squares error (MMSE) technique for the out-of-cell users. The new receiver significantly outperforms the blind receiver and the receiver that ignores the out-of-cell interference. Additionally, due to the nonlinear interference cancellation technique, the proposed receiver can even outperform conventional linear multiuser receivers with the knowledge of all the users.

Browsing by Author "Das, Suman"

Results Per Page

Sort Options