Gerhard Kramer - Research Interests

Multi-User Information Theory

  1. Multi-Hopping for Relay Networks
    Relay networks have a source terminal, one or more relay nodes, and a destination terminal. The source tries to transmit a message to the destination with the help of the relays. The capacity of such networks have been investigated with various restrictions on the nodes.

    • "Cooperative strategies and capacity theorems for relay networks,"
      G. Kramer, M. Gastpar, and P. Gupta, IEEE Trans. Inform. Theory, vol. 51, no. 9, pp. 3037-3063, Sept. 2005.
    • "Models and theory for relay channels with receive constraints,"
      G. Kramer, 42nd Annual Allerton Conf. on Commun., Control and Comp., (Monticello, IL, USA), Sept. 29-Oct. 1, 2004.

  2. Network Coding
    Network coding refers to the combining of packets at the network layer so as to improve the throughput and reliablity of network communication. The impact of network coding on networks with bidirected links was investigated. New edge-cut bounds on network coding rates were developed.

    • "The multicast capacity of deterministic relay networks with no interference,"
      N. Ratnakar and G. Kramer, IEEE Trans. Inform. Theory, vol. 52, no.6, pp. 2425-2432, June 2006.
    • "Edge-cut bounds on network coding rates,"
      G. Kramer and S.A. Savari, Journal of Network and Systems Management, vol. 14, no. 1, pp. 49-67, March 2006.
    • "On networks of two-way channels,"
      G. Kramer and S.A. Savari, in Algebraic Coding Theory and Information Theory, DIMACS Workshop, Dec. 15-18, 2003, Rutgers University, DIMACS Series in Discrete Mathematics and Theoretical Computer Science, vol. 68, A. Ashikhmin and A. Barg, eds., pp. 133-143.

  3. The Discrete Memoryless Network
    A discrete memoryless network (DMN) is a multi-terminal channel that models a wide variety of networks, e.g., multi-access channels, broadcast channels, relay channels, and so on. Bounds on the capacity region were developed, and new capacity points were found.

    • "Capacity results for the discrete memoryless network," G. Kramer, IEEE Trans. Inform. Theory, vol. 49, no.1, pp. 4-21, Jan. 2003.

  4. Capacity of Interference Channels
    The simplest interference channel has four terminals: transmitters 1 and 2 try to send messages to receivers 1 and 2, respectively. The transmitted signals interfere with each other, and this seems to make the problem hard to solve. New outer bounds on the capacity region of Gaussian interference channels were obtained.

    • "Outer bounds on the capacity of Gaussian interference channels," G. Kramer, IEEE Trans. Inform. Theory, vol. 50, no. 3, pp. 581-586, March 2004.

  5. Coding Strategies for Gaussian Interference Networks
    A Gaussian interference network is a channel for which the signals of T transmitters are linearly combined and corrupted by Gaussian noise before reaching R receivers. The receivers feed back their signals to the transmitters to help them code. Feedback strategies were designed that improve the rates that are achievable without feedback. These strategies are optimal for multi-access channels if the signal-to-noise ratio is beyond some threshold.

    • "Feedback strategies for white Gaussian interference networks," G. Kramer, IEEE Trans. Inform. Theory, vol. 48, no. 6, pp. 1423-1438, June 2002.

Channel Coding, Iterative Decoding, and Modulation
  1. Low-Density Parity-Check Codes and Modulation
    A methodology is developed for combining low-density parity-check (LDPC) codes and modulation signal sets. The LDPC codes are designed with extrinsic information transfer (EXIT) charts, and are shown to perform well for multi-antenna communication.

    • "Design of low-density parity-check codes for modulation and detection," S. ten Brink, G. Kramer and A. Ashikhmin, IEEE Trans. Commun., vol. 52, no. 4, pp. 670-678, April 2004.

  2. Repeat-Accumulate Codes for Iterative Detection and Decoding
    Repeat-accumulate (RA) codes are designed for multi-antenna communication. These RA codes perform just as well as LDPC codes.

    • "Design of repeat-accumulate codes for iterative detection and decoding," S. ten Brink and G. Kramer, IEEE Trans. Signal Proc., vol. 51, no. 11, pp. 2764-2772, Nov. 2003.
    • "Turbo processing for scalar and vector channels," S. ten Brink and G. Kramer, Proc. 3rd Int. Symp. on Turbo Codes, (Brest, France), pp. 23-30, Sept. 1-5, 2003.

  3. Decoding Model and Properties of Extrinsic Information Transfer (EXIT) Charts
    EXIT charts are a tool for predicting the convergence behavior of iterative decoding and detection schemes. Several properties of EXIT charts are found that are useful for guiding code design.

    • "Extrinsic information transfer functions: model and erasure channel properties," A. Ashikhmin, G. Kramer, and Stephan ten Brink, IEEE Trans. Inform. Theory, vol. 50, no. 11, pp. 2657-2673, Nov. 2004.

  4. Spectral Efficiency of Coded Phase Shift Keying for Fiber Optic Communication
    Differentially detected phase shift keying (D-PSK) gives substantial energy gains over direct-detection on-off keying. Several coded modulation schemes based on D-PSK were evaluated.

    • "Spectral efficiency of coded phase-shift keying for fiber-optic communication," G. Kramer, A. Ashikhmin, A.J. van Wijngaarden, and X. Wei, IEEE/OSA J. Lightwave Techn., vol. 21, no. 10, pp. 2438-2445, Oct. 2003.

Source Coding
  1. Multiple-Description Coding
    The problem of multiple description (MD) coding is to transmit a source stream to a destination via several channels. The channels are modeled as being either error-free or completely noisy. New information-theoretic inner and outer bounds on the set of achievable rates and distortions were developed.

    • "Multiple description coding with many channels," R. Venkataramani, G. Kramer and V.K. Goyal, IEEE Trans. Inform. Theory, vol. 49, no. 9, pp. 2106-2114, Sept. 2003.

  2. Quantum Data Compression
    A quantum source is a device that emits "qubits" represented by density operators. The case when the set of density operators commute was recently reformulated as a classical information theory problem. This classical problem was solved.

    • "Communicating probability distributions," G. Kramer and S.A. Savari, IEEE Trans. Inform. Theory, to appear Feb. 2007.


Last modified: Dec. 2006