PaperlayerView exact revision on arXiv

Source-linked AI summary

Massive MIMO for Next Generation Wireless Systems

Erik G. Larsson, Ove Edfors, Fredrik Tufvesson, Thomas L. Marzetta

arXiv:1304.6690v3cs.IT

TL;DR

Massive MIMO promises scalable wireless capacity and efficiency, but limited pilot resources and pilot contamination remain important performance concerns. This paper surveys the concept and recent research, reporting potential capacity gains of 10 times or more and radiated energy-efficiency gains of about 100 times.

  • Problem

    Massive MIMO must scale despite limited orthogonal pilot sequences and pilot contamination that may limit performance as antenna numbers increase.

  • Method

    The paper provides an overview of massive MIMO and synthesizes recent work on energy efficiency, excess degrees of freedom, TDD calibration, pilot contamination, and channel measurements.

  • Results

    Massive MIMO can increase capacity 10 times or more while improving radiated energy-efficiency by about 100 times.

  • Takeaways & Limitations

    Massive MIMO has substantial potential as a beyond-4G technology using parallel low-cost, low-power units at base stations and mobile terminals.

  • Takeaways & Limitations

    Pilot contamination appears likely to require mitigation and may constitute an ultimate performance limit under some assumptions and pilot-based channel estimation.

Abstract

from arXiv · show

Multi-user Multiple-Input Multiple-Output (MIMO) offers big advantages over conventional point-to-point MIMO: it works with cheap single-antenna terminals, a rich scattering environment is not required, and resource allocation is simplified because every active terminal utilizes all of the time-frequency bins. However, multi-user MIMO, as originally envisioned with roughly equal numbers of service-antennas and terminals and frequency division duplex operation, is not a scalable technology. Massive MIMO (also known as "Large-Scale Antenna Systems", "Very Large MIMO", "Hyper MIMO", "Full-Dimension MIMO" & "ARGOS") makes a clean break with current practice through the use of a large excess of service-antennas over active terminals and time division duplex operation. Extra antennas help by focusing energy into ever-smaller regions of space to bring huge improvements in throughput and radiated energy efficiency. Other benefits of massive MIMO include the extensive use of inexpensive low-power components, reduced latency, simplification of the media access control (MAC) layer, and robustness to intentional jamming. The anticipated throughput depend on the propagation environment providing asymptotically orthogonal channels to the terminals, but so far experiments have not disclosed any limitations in this regard. While massive MIMO renders many traditional research problems irrelevant, it uncovers entirely new problems that urgently need attention: the challenge of making many low-cost low-precision components that work effectively together, acquisition and synchronization for newly-joined terminals, the exploitation of extra degrees of freedom provided by the excess of service-antennas, reducing internal power consumption to achieve total energy efficiency reductions, and finding new deployment scenarios. This paper presents an overview of the massive MIMO concept and contemporary research.

1 Background: Multi-User MIMO Maturing

MU-MIMO uses multiple antennas to transmit multiple data streams to several terminals simultaneously, improving data rate, reliability, energy efficiency, and interference management. Conventional MU-MIMO is maturing and has been incorporated into 4G LTE and LTE-Advanced standards, although the benefits cannot all be achieved simultaneously and depend on propagation conditions.

  • 1 Background: Multi-User MIMO Maturing: MU-MIMO communicates with several terminals simultaneously using multiple antennas to transmit multiple data streams.MIMO relies on multiple antennas, while the multiuser form serves several terminals at the same time.
  • 1 Background: Multi-User MIMO Maturing: More antennas can increase data rate by enabling more independent streams and simultaneously serving more terminals.The passage attributes this improvement directly to the number of antennas.
  • 1 Background: Multi-User MIMO Maturing: More antennas can enhance reliability through a greater number of distinct propagation paths.Reliability improves because signals can propagate over more distinct paths.
  • 1 Background: Multi-User MIMO Maturing: MU-MIMO can improve energy efficiency by focusing emitted energy toward terminals and reduce interference by avoiding harmful transmission directions.These gains arise from spatially directing energy and deliberately suppressing transmission where interference would be harmful.
  • 1 Background: Multi-User MIMO Maturing: Conventional MU-MIMO is maturing and has been incorporated into recent and evolving 4G LTE and LTE-Advanced standards.The passage identifies LTE and LTE-A as examples of wireless broadband standards incorporating MU-MIMO.
  • 1 Background: Multi-User MIMO Maturing: The four benefits cannot all be achieved simultaneously, and their realization depends on propagation conditions.The passage presents data rate, reliability, energy efficiency, and interference reduction as general benefits subject to these constraints.

2 Going Large: Massive MIMO

Massive MIMO scales MIMO to arrays of a few hundred antennas serving many tens of terminals in the same time-frequency resource. The section presents its energy-efficient, secure, robust, spectrum-efficient promise and focuses on key implementation and measurement challenges.

  • Research focus: Current research emphasizes energy efficiency, excess degrees of freedom, TDD calibration, pilot-contamination mitigation, and new channel measurements.The paper follows earlier exposition by focusing on developments from the last three years.
  • System concept: Massive MIMO uses a few hundred antennas to simultaneously serve many tens of terminals in the same time-frequency resource.Its basic premise is to reap conventional MIMO benefits on a much greater scale.
  • System concept: Massive MIMO is positioned as an enabler of future broadband networks that are energy-efficient, secure, robust, and spectrum-efficient.The section frames these properties as overall anticipated benefits of the technology.
  • Channel knowledge: Spatial multiplexing requires sufficiently accurate uplink and downlink channel knowledge, obtained through terminal pilots on the uplink.Downlink channel acquisition is more difficult than uplink acquisition in conventional MIMO systems.
  • Experimental status: Although massive MIMO has been mostly theoretical, basic testbeds and initial channel measurements are becoming available.The section notes that the concept has stimulated research in random matrix theory and related mathematics.

3 The Potential of Massive MIMO

Massive MIMO combines phase-coherent, computationally simple processing across many base-station antennas with aggressive spatial multiplexing. This enables major capacity and radiated-energy-efficiency gains while also supporting inexpensive hardware, lower latency, simpler multiple access, and greater interference robustness.

  • Capacity and energy efficiency: 10 times or more capacity and radiated energy-efficiency in the order of 100 times are identified as key massive MIMO gains.These gains are attributed to aggressive spatial multiplexing and sharply focused energy from many antennas.
  • Signal processing: Maximum-ratio combining is attractive because it is computationally simple and can be performed independently at each antenna unit.Its effectiveness in massive MIMO follows from channel responses for different terminals tending to become nearly orthogonal as the antenna count grows.
  • Capacity and energy efficiency: Serving many tens of terminals simultaneously in the same time-frequency resource can make spectral efficiency 10 times higher than conventional MIMO.With MRC and power scaled down without seriously affecting overall spectral efficiency, multiuser interference and hardware imperfections tend to be overwhelmed by thermal noise.
  • Implementation: Massive MIMO replaces expensive ultra-linear 50 Watt amplifiers with hundreds of low-cost amplifiers operating in the milli-Watt range.This architecture can also eliminate expensive and bulky items such as large coaxial cables, while relaxing accuracy and linearity requirements for individual RF chains.
  • Energy and deployment: Two orders of magnitude less total output RF power can enable operation from wind or solar and substantially reduce emitted electromagnetic interference.The lower power consumption is presented as important for cellular base stations and deployments without an electricity grid.
  • System benefits: Massive MIMO significantly reduces air-interface latency, simplifies the multiple-access layer, and improves robustness to unintended interference and intentional jamming.Intentional jamming is described as a growing cybersecurity concern affecting civilian wireless systems, including public-safety applications.

4 Limiting Factors of Massive MIMO

The main limiting factors are hardware non-reciprocity in TDD operation, scarce orthogonal pilots and pilot contamination, and propagation conditions that can vary across large arrays. Measurements nevertheless provide compelling evidence that favorable propagation is substantially valid in practice, with about 10 times more base-station antennas than users yielding stable performance not far from ideal.

  • Reciprocity and calibration: TDD depends on channel reciprocity, but uplink and downlink hardware chains may be non-reciprocal; calibration-based solutions do not appear to be a serious problem.The propagation channel itself is essentially reciprocal, while transceiver hardware chains may differ between uplink and downlink.
  • Reciprocity and calibration: Full beamforming gains do not require terminal-chain calibration when the base-station equipment is properly calibrated.Residual terminal receiver mismatch can be handled with supplementary pilots whose overhead is very small.
  • Pilot contamination: About 200 orthogonal pilot sequences are available in a typical one millisecond coherence interval, so the supply is easily exhausted as more terminals are served.The maximum number of orthogonal pilots is upperbounded by coherence-interval duration divided by channel delay-spread.
  • Pilot contamination: Pilot contamination creates channel-estimation errors and directed uplink and downlink interference that grows with the number of service-antennas at the same rate as the desired signal.Reusing pilots across cells contaminates an estimated channel with a linear combination of channels from terminals sharing that pilot.
  • Pilot contamination: Pilot contamination may constitute an ultimate performance limit as antenna count grows without bound for pilot-based receivers, although this conclusion has been contested under specific assumptions.Pilot contamination affects massive MIMO more profoundly than classical MIMO, while pilot allocation, blind estimation and network-aware precoding are proposed mitigation directions.
  • Favorable propagation: About 10 times more base-station antennas than users appears sufficient for stable performance not far from ideal, despite array configuration and propagation-dependent convergence.Large arrays can exhibit fading and changing small-scale statistics across the aperture; physically large linear arrays approach theoretical i.i.d. performance, while compact circular arrays can perform worse because of smaller aperture.

5 Massive MIMO: a Goldmine of Research Problems

Massive MIMO makes many traditional communication-theory problems less relevant while creating a broad research agenda. Key challenges span scalable processing, low-cost and imperfect hardware, power consumption, channel knowledge, deployment, and validation.

  • Processing: Massive MIMO arrays generate vast baseband data requiring real-time, simple linear or nearly linear processing and optimized algorithms.Downlink precoding offers substantial room for further innovation.
  • Hardware: Building hundreds of RF chains, converters, and related components requires handset-scale economies of manufacturing.The challenge is to achieve low cost at massive hardware scale.
  • Hardware: Low-cost components introduce phase noise, I/Q imbalance, quantization noise, and relaxed-linearity power-amplifier challenges that must be managed.Massive MIMO relies on averaging noise, fading, and some interference, but hardware imperfections may be larger in practice.
  • Energy efficiency: Reducing radiated power by a thousand times is insufficient unless baseband processing and other internal consumption are also addressed.Highly parallel or dedicated baseband hardware is identified as an important research direction.
  • Channels and TDD: Research must improve channel models, reciprocity calibration, and pilot-contamination mitigation to enable realistic assessment and reliable TDD operation.Calibration raises questions about frequency and time resources and additional hardware, while pilot contamination may be more limiting than in traditional MIMO.
  • Deployment and validation: New deployment scenarios, comparative system studies, and prototypes are needed to establish massive MIMO’s role alongside small-cell and HetNet approaches.The Argos testbed demonstrated feasibility with 64 coherently operating antennas and TDD based on channel reciprocity.

6 Conclusions and Outlook

Massive MIMO is presented as a promising enabling technology for beyond-4G cellular systems, offering gains in efficiency, robustness, reliability, and hardware affordability. These benefits also motivate substantial new research in academia and industry.

  • Conclusions and Outlook: Massive MIMO has large potential as an enabling technology for future beyond-4G cellular systems.The paper frames the technology as a key candidate for future cellular networks.
  • Conclusions and Outlook: The technology offers advantages in energy efficiency, spectral efficiency, robustness, and reliability.These benefits are identified as central strengths of massive MIMO systems.
  • Conclusions and Outlook: Massive MIMO enables the use of low-cost hardware at both base stations and mobile units.The affordability benefit applies to both sides of the wireless link.
  • Conclusions and Outlook: The technology creates many new research problems for academia and industry.The paper characterizes these unresolved problems as a substantial research opportunity.

Biographies

The paper’s authors are professors and researchers in communications, signal processing, propagation, and wireless systems, with extensive academic and industry experience. Thomas L. Marzetta is identified as an early proponent of Massive MIMO and its efficiency gains over 4G technologies.

  • Erik G. Larsson: Erik G. Larsson is Professor and Head of Communication Systems at Linköping University, with about 100 journal papers and a textbook co-authorship.He is also an Associate Editor for IEEE Transactions on Communications and received the IEEE Signal Processing Magazine Best Column Award in 2012.
  • Ove Edfors: Ove Edfors is Professor of Radio Systems at Lund University, researching statistical signal processing, low-complexity algorithms, and Massive MIMO system performance.His Massive MIMO work examines how realistic propagation characteristics affect performance and base-band processing complexity.
  • Fredrik Tufvesson: Fredrik Tufvesson is an associate professor at Lund University whose research includes channel measurements and modeling for MIMO and UWB wireless communications.He also works on distributed antenna systems and radio-based positioning, after receiving his Ph.D. from Lund University in 2000 and spending almost two years at Fiberless Society.
  • Thomas L. Marzetta: Thomas L. Marzetta received his PhD from MIT, joined Bell Laboratories in 1995, and was an early proponent of Massive MIMO’s efficiency improvements over 4G technologies.He later directed the Communications and Statistical Sciences Department within Bell Laboratories’ former Mathematical Sciences Research Center.
Loading 1304.6690v3…