Skip to Main Content
Communication at millimeter wave (mmWave) frequencies is defining a new era of wireless communication. The mmWave band relieves spectral gridlock at lower frequencies by offering much higher bandwidth communication channels than presently used in commercial wireless systems. The next generation of wireless local area networks is exploiting the mmWave unlicensed band at 60 GHz to provide multigigabit-per-second data rates. There is also growing interest in using mmWave licensed spectrum for 5G cellular systems. MmWave communication could also provide important benefits in other application scenarios like wearable networks, vehicular communications, or autonomous robots. The potential for mmWave is immense.
Signal processing is critical for enabling the next generation of mmWave communication. Because of the wide bandwidth, overall complexity and mixed signal power consumption are significant concerns. This motivates developing MIMO signal processing techniques, e.g., that have to operate with few high-resolution or many low-resolution analog-to-digital converters. The propagation channel characteristics lead to sparsity in the channel, which can be exploited in channel estimation, signal detection, precoder/combiner design, and equalization. System analysis of mmWave wireless systems is more complicated due to the use of compact antennas, sensitivity to blockages, and distance dependent propagation effects. Relays may play an important role in mmWave to provide inband backhaul for cellular networks or to enhance coverage in the presence of blockages. Because of the higher carrier frequencies, supporting mobility becomes a significant challenge, requiring the development of time-varying signal processing techniques such as rapid beam adaptation.
This special issue brings together contributions from researchers and practitioners in the area of signal processing for wireless communications with an emphasis on communication at millimeter wave frequencies. In the end, eleven papers were selected for inclusion in the special issue.
This issue starts off with an overview paper written by the guest editors entitled “An Overview of Signal Processing Techniques for Millimeter Wave MIMO Systems.” It provides an overview on topics of interest for signal processing researchers including propagation and channel models, MIMO architectures for mmWave, precoding and combining techniques for mmWave MIMO systems, and channel estimation exploiting sparsity. This paper provides foundations for the other contributions in this issue.
The second paper entitled “Proposal on Millimeter-Wave Channel Modeling for 5G Cellular System,” proposes a mmWave channel model for 5G cellular systems, inspired by several measurement campaigns. The core idea is to supplement existing measurement results with data obtained using ray tracing. Parameters are provided to generate the channel coefficients using a ray-based frequency selective channel. The model in this paper can be used to study mmWave performance in urban areas.
The next paper entitled “Analog multiband: efficient bandwidth scaling for mm-wave communication,” presents an architecture for wideband mmWave communication systems. The idea is to provide an efficient means for channelizing a wideband signal into multiple subchannels that can be processed in parallel. The impact due to imperfect channelization is quantified and techniques for mitigating it are proposed. The results in this paper provide a potential solution for hardware efficient wideband operation in mmWave.
The next two papers deal with hybrid analog/digital precoding, which is a power-efficient MIMO architecture. The first paper is entitled “Alternating Minimization Algorithms for Hybrid Precoding in Millimeter Wave MIMO Systems.” This paper provides algorithms that alternate between the design of the analog and digital precoders. Both fully connected and partially connected array mappings are assumed. The performance gap between the different algorithms and the fully digital solution is quantified in simulations. The second paper is entitled “Hybrid Digital and Analog Beamforming Design for Large-Scale Antenna Arrays.” This paper shows that with twice as many RF chains as desired data streams and unquantized phase shifters in the analog precoding network, there is no loss in the hybrid architecture compared to the fully digital solution. It then proposes algorithms for designing precoders for the case where zero loss is not guaranteed. Both papers on hybrid precoding provide new algorithms that confirm the viability of the hybrid precoding solution.
Next, this issue contains two papers on estimation of mmWave channels. The first paper entitled “Compressive channel estimation and tracking for large arrays in mm wave picocells,” proposes a strategy which exploits sparsity in the channel. The proposed technique involves a specially designed training beacons and feedback from the mobile stations to allow the base station to estimate the path gains and angles-of-departure for all the users at the same time. Refinement and tracking are also included. The second paper is entitled “Subspace Estimation and Decomposition for Large Millimeter-Wave MIMO Systems.” This paper exploits reciprocity to develop a method for estimating subspaces of the channel using channel reciprocity and exploiting sparsity of the eigenmodes. Then, an iterative algorithm that accounts for the hybrid analog/digital structure is proposed. Results show that the gap is close to the fully digital solution at medium-to-high SNR. Together, these papers show that channel estimation, while different in mmWave systems, is practically feasible.
The next three papers address problems related to mmWave networks. The first paper entitled “Beamforming Tradeoffs for Initial UE Discovery in Millimeter-Wave MIMO Systems” deals with the topic of initial user discovery. It studies the problem of learning the right singular vector in a system with only analog beamforming. The paper also suggests a broadcast-based solution using limited feedback with specially designed directional codebooks. Comparisons are made between different beamforming strategies in terms of different metrics. The second paper is entitled “On the Performance of Random Beamforming in Sparse Millimeter Wave Channels.” This paper analyzes the performance of random beamforming in a channel called the uniform random multipath model. Results are provided, e.g., on how the number of users should scale to achieve linear sum rate scaling. Several other extensions are also provided to different kinds of beamforming. The third paper is entitled “On the Performance of Relay Aided Millimeter Wave Networks.” This paper shows that how relays can improve performance in mmWave networks. Using a stochastic geometry framework, this paper provides results on the end-to-end SNR and suggests relay selection techniques that achieve good performance. These mmWave networking papers show that mmWave has promise not just in single links, but also in networks.
The final paper is entitled “Feasibility of Mobile Cellular Communications at Millimeter Wave Frequency.” This paper describes a design for a radio frame structure to support mmWave communication. The proposed structure was tested as part of a prototype that delivered gigabits-per-second to both static and mobile users in different scenarios. Overall, this paper provides perspective that mmWave is practical and can meet critical performance needs in cellular systems.
We received many papers in response to the call for papers of this special issue. Based on relevance and fit for the special issue, many high-quality papers could not be included. We would like to thank all the authors who submitted their manuscripts to this issue. We would also like to thank the reviewers who lent their precious time to evaluate papers for the issue. Finally, we hope that the wide range of papers in this special issue spurs the future development of heterogeneous networking solutions within the signal processing community.
Back to Top