Research Projects

Modulation and decoding in digital transmission systems

List of researchers

• Danyo Danev , associate professor, Communication Systems, ISY.
• Daniel Puaca , Ph.D. student, Communication Systems, ISY.

Background and industrial relevance

The need for faster and more reliable communication over different transmission media has continued to rise in resent years. With the adoption of the 3G standard within the mobile phone communications a step towards meeting these quality requirements has been made. However the data transmission speed achievable in the system defined by this standard is far from the speed that we are acquainted with while using the modern Internet. In our opinion the need for more flexible access to Internet will be a key issue in the development of the next generation mobile networks. This will impose stronger requirements on the wireless communication links where key layers are the modulation and the error-correction coding and decoding.

In resent talks with researchers from the telecommunication company Ericsson in Linköping it was suggested that block coding will be used for increasing the reliability in the next generation of the mobile telephony standard. A close cooperation for research in this area will be initiated soon. This will be a part of an alliance between research groups and industrial companies with interests in the field of telecommunication.

Project plan and goal

An important part in every communication link is the representation of the digital information in a form suitable for transmission over the physical channel. This representation is known as modulation. Various modulation techniques are known and frequently used in communication links today. They can be represented as constellations of points in Euclidean vector spaces. Good arrangements provide reliable, to some extend, transmission even in the presence of channel imperfections. Another important issue when choosing the modulation scheme is the easiness and the effectiveness of the information symbol retrieval. However, the modulation alone is not able to provide the reliability level needed for the link. That is why certain error-correction coding is introduced prior to the modulation step. In this procedure some additional information is added to the (possibly compressed and encrypted) information sequence in order to recover some of the errors caused by the noise in the channel. The task of the decoding step at the receiver end is to reconstruct the uncoded information block as fast as possible. Another condition for the decoding is to accurately correct as many error patterns as possible.

The main goal of the project is to investigate the relationship between the modulation techniques used in information transmission and the decoding algorithms used in connection to the detector at the receiver end. Presentation and analysis of many decoding schemes are published in scientific journals. However, most of their properties are given in comparison with other decoding techniques while the modulation is preserved. Our intention is to look closer on the influence of the type of modulation used on the performance of a given decoding algorithm.

The projects main task during the first year will be to investigate existing combinations of modulation and decoding. It should be noted that this investigation includes wireless as well as wire bound communication schemes. The reason for this is that the reliability of the channel is a function of the distance between the different points chosen in the Euclidean vector space. However, different modulation techniques will yield a difference in performance depending on whether they are used for wireless or wire bound communication.

In the modulation schemes that are to be considered, the initial investigation will include an analysis of certain known, and well used, error-correcting techniques and the impact of different decoding techniques on their performance. The importance of efficient decoding is even greater when the communication link is a wireless one. This is due to the rapidly changing condition of the transmission channel. As an obvious example one could mention a cellular phone user who is moving while using a high data rate service. In such varying channels there are two main approaches for maintaining reliable communication. One is to have a flexible code which can be adapted to changing link performance, and thus it will be increasing the coding gain at the expense of the amount of information being transferred. The second approach is to vary the modulation scheme, for instance by using only a subset of signals, which is equivalent of considering only a subset of points in the Euclidean vector space, and thereby also increasing the reliability while decreasing the information rate. Of course one method doesn't exclude the other, and combinations of both are to be considered. One should note that the use of codes with higher coding gains induce stronger demands on the decoding to fully take advantage of the redundant information introduced.

It can also be noted that many modulation techniques don't use strict Gray coding, and as a consequence there could differ more than one bit between adjacent points in the Euclidean space. In such communication links the need for proper coding and, above all, decoding is of even greater importance.

All these cases will be part of the initial investigation. As a reference the performance of other methods than block-coding will also be investigated, as of today they are widely used.

Project visions

The long term direction of this project will be to further strengthen the positions of the Data transmission group as one of the leaders in the field of coding theory on a national as well as international scene. This project is aimed to bring the research within the group closer to problems arising from the development of practical applications. Another goal is to bring the cooperation between the research groups at the University and Ericsson to a higher level. We truly hope that our research will be considered when the standardisation process for the next generation of mobile telephony starts.

In a three years therm this project should result in at least one PhD thesis as well as several published research papers.

Relevance to other CENIIT financed projects

There is no direct connection to a project currently supported by CENIIT. The following projects share common interests in the field of telecommunications with the project described here.

• Konstruktion och implementering av digitala signalbehandlande system, Håkan Johansson, ISY
• Optimal design och effektiv planering av telekommunikationssystem, Peter Värbrand, ITN
• Flexible frequency band reallocation, Per Löwenborg, ISY

Publications

• D Puaca: A modification to the Reed decoding algorithm
• M. Tjader, M. Grimnell, D. Danev and H.M. Tullberg: Efficient Message-Passing Decoding of LDPC Codes using Vector-based Messages
• D. Danev and S. Dodunekov: A Family of Ternary Quasi-Perfect Codes
• D. Danev, and D. Puaca: New majority logic decoding algorithm for Reed-Muller codes