PaperlayerView exact revision on arXiv

Source-linked AI summary

A Survey on Spatial Modulation in Emerging Wireless Systems: Research Progresses and Applications

Miaowen Wen, Beixiong Zheng, Kyeong Jin Kim, Marco Di Renzo, Theodoros A. Tsiftsis, Kwang-Cheng Chen, Naofal Al-Dhahir

arXiv:1907.02941v1cs.IT

TL;DR

Spatial modulation seeks to improve energy and spectral efficiency while reducing RF-chain hardware costs in emerging wireless systems. This survey synthesizes SM principles, variants, enhancements, integrations, applications, and extensions, concluding that SM offers a useful compromise across diverse implementations.

  • Problem

    Emerging wireless systems need modulation approaches that balance spectral efficiency, energy efficiency, deployment cost, and hardware complexity.

  • Method

    The survey reviews SM principles, system variants, performance enhancements, combinations with other techniques, applications, and extensions to additional signal domains.

  • Results

    SM uses transmit-antenna activation states to convey additional information while requiring fewer RF chains, supporting an attractive spectral–energy-efficiency compromise.

  • Takeaways & Limitations

    SM has broad prospects across emerging wireless implementations, including mmWave systems, physical-layer security, compressed-sensing detection, and generalized activation designs.

Abstract

from arXiv · show

Spatial modulation (SM) is an innovative and promising digital modulation technology that strikes an appealing trade-off between spectral efficiency and energy efficiency with a simple design philosophy. SM enjoys plenty of benefits and shows great potential to fulfill the requirements of future wireless communications. The key idea behind SM is to convey additional information typically through the ON/OFF states of transmit antennas and simultaneously save the implementation cost by reducing the number of radio frequency chains. As a result, the SM concept can have widespread effects on diverse applications and can be applied in other signal domains such as frequency/time/code/angle domain or even across multiple domains. This survey provides a comprehensive overview of the latest results and progresses in SM research. Specifically, the fundamental principles, variants of system design, and enhancements of SM are described in detail. Furthermore, the integration of the SM family with other promising techniques, applications to emerging communication systems, and extensions to new signal domains are also extensively studied.

I. INTRODUCTION

Spatial modulation addresses growing wireless demands by encoding information in transmit-antenna activation while reducing RF-chain requirements. The survey reviews SM principles, variants, enhancements, integrations, applications, and extensions across signal domains.

  • Future wireless systems face demands for ultra-high capacity, ultra-low latency, and massive connectivity amid scarce spectrum and rising power consumption.
  • SM offers benefits including higher energy efficiency, lower detection and RF-circuit complexity, freedom from inter-channel interference, and compatibility with massive MIMO.
  • SM conveys extra information through the index of one active transmit antenna alongside conventional constellation symbols, requiring only one RF chain for antenna activation.This reduces energy consumption and hardware cost while avoiding inter-channel interference and inter-antenna synchronization requirements.
  • Receivers detect active-antenna indices and demodulate constellation symbols, but jointly achieving effective detection and low complexity remains challenging across channel conditions.
  • SM can combine with compressed sensing, mmWave MIMO, NOMA, SWIPT, physical-layer security, and visible-light communications for emerging wireless applications.Inactive antennas can support energy harvesting in SWIPT, while antenna switching can create fast time-varying environments that degrade eavesdropper decoding.
  • The survey covers SM fundamentals, variants, performance enhancements, integrations, emerging-system applications, extensions to other domains, and future research directions.

II. BASIC PRINCIPLE AND VARIANTS OF SM

SM conveys information through both constellation symbols and active-antenna indices, typically using a single RF chain. Its variants and receiver designs trade spectral efficiency, hardware complexity, detection complexity, and performance.

  • Single-RF SM: SM selects one active transmit antenna to carry a constellation symbol, adding spatial-index information while requiring only one RF chain.The antenna index changes according to incoming information bits, while the selected antenna transmits the constellation symbol.
  • Variants: When M = 1, SM becomes SSK, mapping all information bits to the active-antenna index rather than a constellation symbol.SSK therefore represents the spatial-index-only specialization of SM.
  • Detection: The SM receiver detects both the active-antenna index and the constellation symbol using joint or decoupled detection.Maximum-likelihood detection searches jointly over antennas and symbols, whereas decoupled detection separates these tasks.
  • Detection: Decoupled detection reduces the search space from O(NT ∗M) to O(NT + M), but usually degrades performance relative to optimal detection.This creates a central design challenge: achieving low complexity without sacrificing detection performance.
  • Extensions: SM can support full-duplex nodes by selecting antennas for transmission and reception simultaneously, increasing spectral efficiency while retaining RF-chain advantages.When NT is not a power of 2, fractional bit encoding, bit padding, or varying constellation order can improve spectral efficiency, with increased error-propagation susceptibility.

B. Generalized (G)SM

Generalized spatial modulation ((G)SM) relaxes single-antenna activation by selecting K of N_T antennas, improving spatial information while retaining one-RF-chain operation. It flexibly trades spectral efficiency, deployment cost, and error performance, but detection complexity grows with the search space.

  • System design: (G)SM selects K out of N_T transmit antennas to carry the same PSK/QAM symbol, while inactive antennas remain off.The scheme preserves one-RF-chain operation, avoids inter-channel interference, and requires no inter-antenna synchronization.
  • Spectral efficiency: Allowing different active antennas to carry different PSK/QAM symbols further increases spectral efficiency to include a K log2 M contribution.This extends the version where all active antennas transmit the same constellation symbol.
  • Design trade-offs: (G)SM provides a compromise among spectral efficiency, deployment cost, and error performance by choosing the number of active antennas.Single-RF SM and full-activation MIMO correspond to K = 1 and K = N_T, respectively.
  • System design: Incoming bits split into spatial and constellation parts: one selects legal active-antenna combinations, while the other generates symbols for transmission.The legal combinations are indexed through lookup tables or combination strategies, and the transmit vector has K nonzero entries.
  • Detection: The receiver jointly detects active-antenna indices and constellation symbols from the received MIMO signal.The signal model uses the channel columns associated with the selected active antennas, transmitted symbols, power scaling, and AWGN.
  • Detection: The ML detector searches O(N_L · M^K) possibilities, creating high complexity for large-scale MIMO or high-order constellations.Low-complexity detectors may be suboptimal or channel-dependent; DSMC and K-best sphere decoding target near-optimal performance at lower complexity, with Fig. 3 comparing representative uncoded detectors.

C. Differential (D)SM

Differential spatial modulation (D)SM conveys information through antenna activation patterns across space-time blocks, enabling non-coherent detection without transmitter pilots or receiver channel estimation. Its reported BER loss versus coherent SM is below 3 dB at 3 bps/Hz, while performance improves with more receive antennas.

  • Differential (D)SM: Differential encoding avoids the receiver CSI requirement associated with detecting channel-dependent spatial information.This removes pilot insertion at the transmitter and channel estimation at the receiver.
  • Differential (D)SM: D(SM) uses antenna activation order to select one of N_T! permutation-based space-time blocks from incoming spatial bits and the previous block.The resulting block has one non-zero entry in each row and column.
  • Differential (D)SM: At 3 bps/Hz with N_T = 4, non-coherent (D)SM has less than 3 dB BER performance loss relative to coherent-detection SM.The comparison is reported in Fig. 6.
  • Differential (D)SM: With 3 bps/Hz transmission, the performance gain of (D)SM over single-antenna DPSK increases as the number of receive antennas grows.Fig. 7 compares N_T = 3 and N_T = 4 settings and identifies QPSK and 8PSK in its legend.
  • Differential (D)SM: Subsequent work develops reduced-complexity near-optimal detection, APSK-based (D)SM, and differential schemes using coding, diversity, and algebraic constructions.Applications include dual-hop networks, Gray coding, cyclic constellations, and algebraic field extensions.

D. Receive (R)SM

Receive spatial modulation (RSM) conveys spatial information through receive-antenna indices while transmitter precoding supports beamforming and simplified reception. Research extends RSM across CSI assumptions, multiple active receive antennas, power allocation, and shadowing channels.

  • Receive (R)SM: RSM applies spatial modulation at the receiver, using receive-antenna indices alongside PSK/QAM constellation information.Transmitter precoding provides beamforming gain while retaining a low-complexity design.
  • Receive (R)SM: Initial RSM studies consider zero-forcing and MMSE precoding with perfect transmitter CSI, followed by schemes for imperfect CSI.Generalizations activate more than one receive antenna and investigate error performance and low-complexity detectors.
  • Receive (R)SM: Later analyses derive MMSE-RSM error bounds and study power allocation, diversity order, coding gain, and error performance under shadowing MIMO channels.Approximate power-allocation solutions are proposed for higher performance gain.

III. PERFORMANCE ENHANCEMENT FOR SM

SM performance depends strongly on how distinct the channel signatures associated with different transmit antennas are. Because rich scattering may be unrealistic, preprocessing methods are used to improve performance in adverse propagation environments.

  • Performance Enhancement for SM: SM and its variants rely on distinct channel signatures or fingerprints associated with different transmit antennas.This dependence is identified as a central determinant of system performance.
  • Performance Enhancement for SM: Rich scattering is required to avoid significant performance degradation, but may not be realistic in practical systems.Link adaptation, precoding or TCM, and space-time-coded transmission are described as remedies for adverse environments.

A. Link-Adaptive SM

Link-adaptive SM adjusts transmission parameters or precoding according to channel conditions to improve BER performance and channel utilization. Related schemes use feedback, Huffman mapping, or spatial correlation information for adaptive operation.

  • Link-Adaptive SM: Under slow fading, link-adaptive SM selects modulation order, transmit power, transmit-antenna number, or precoding based on channel conditions.The receiver determines optimal settings after channel estimation.
  • Link-Adaptive SM: Adaptive SM in cognitive radio uses limited reliable feedback to vary modulation type or transmit power for secondary-system spectral and energy efficiency.A separate scheme uses Huffman mapping with variable-length codes to activate antennas with distinct probabilities.

B. Precoding/TCM Aided SM

Precoding and trellis-coded modulation address SM’s vulnerability to spatial correlation, while diversity and compressed-sensing methods improve robustness or detection complexity. These techniques extend SM beyond its basic single-antenna transmission model.

  • Precoding/TCM: Spatial-information detection degrades in spatially correlated fading because antenna channel responses become similar.
  • Precoding/TCM: Precoding can reduce asymptotic average BER while preserving the average transmit-power budget without explicit channel knowledge.
  • Precoding/TCM: TCM partitions transmit antennas into subsets to maximize spatial distance, improving conventional SM performance in both correlated and uncorrelated fading channels.
  • Diversity enhancement: SM lacks transmit-diversity gain, motivating STBC integrations such as Alamouti-coded SM, which achieves second-order transmit diversity with linear-complexity optimal detection.
  • Low-complexity detection: Compressed-sensing detectors exploit sparse (G)SM symbols to reduce complexity, including large-scale MIMO, multiuser, grouped, and STBC-based systems.

B. Non-orthogonal Multiple Access (NOMA) Aided SM

NOMA-aided SM combines spatial indexing with power-domain multiplexing to support connectivity and spectral efficiency, while addressing interference, security, and receiver-complexity concerns through varied designs.

  • NOMA integration: NOMA multiplexes multiple users through superposition coding in the power domain to support massive connectivity and low transmission latency.
  • NOMA integration: NOMA-aided (G)SM improves spectral and energy efficiency by jointly detecting user activity and sparse spatial-modulation signals.
  • NOMA integration: Switching or hybrid NOMA-SM schemes address performance degradation when users have similar channel gains by estimating antenna indices and constellation symbols separately.
  • Low-complexity designs: Antenna partitioning can avoid successive-interference cancellation and multiuser interference, while cooperative designs separately deliver spatial-index and constellation information.
  • Security provisioning: SM’s antenna switching and wireless-channel randomness support physical-layer security by making spatial information harder for eavesdroppers to recover.
  • Security provisioning: Precoding, antenna selection, mapping variation, and cooperative relaying are investigated to improve secrecy rates or protect spatial and constellation information.

V. APPLICATIONS OF SM TO EMERGING COMMUNICATION SYSTEMS

In mmWave systems, (G)SM uses spatial indexing and a reduced number of RF chains alongside analog precoding. This design lowers implementation cost while retaining precoding gains and supporting high-frequency communications.

  • A. SM in mmWave Communications: mmWave communications offer high data services but require beamforming to compensate for high propagation loss at around 60 GHz.
  • A. SM in mmWave Communications: (G)SM addresses the prohibitively costly RF-chain requirements of compact large-antenna mmWave MIMO by activating antenna subsets connected to fewer RF chains.
  • Transmitter architecture: The transmitter separates incoming data into constellation and spatial domains, feeding constellation symbols through RF chains and antenna groups through analog precoders.
  • Transmitter architecture: When NRF = 1, the architecture becomes single-RF SM, providing extremely low implementation and detection cost in mmWave communications.
  • System benefits: (G)SM-aided mmWave communication reduces transmitter cost through fewer RF chains while obtaining analog-precoding gains.
  • System benefits: Reduced-RF-chain (G)SM can approach the constrained capacity of full-RF spatial multiplexing, while variable activation and antenna separation are also studied for efficiency.

B. SM in Optical Wireless Communications (OWC)

SM extends to optical wireless, hardware-constrained, and simultaneous information-and-power-transfer systems. The surveyed designs target efficient transmission, practical deployment, and energy harvesting across diverse implementations.

  • OWC context: Optical wireless communications complement radio systems by providing high-rate transmission, including energy-efficient indoor lighting and communication with LEDs.
  • Optical SM: Optical SM has been developed for indoor and outdoor links, with coded and uncoded performance analyzed under turbulence-induced fading.
  • Optical SM: VLC applies SM in RF-free environments such as hospitals, airplanes, and gas stations, using fast-switching LED arrays and MIMO to support high rates.
  • Hardware implementation: Most SM studies assume ideal transceiver hardware, whereas channel-estimation errors, motion, phase noise, I/Q imbalance, and amplifier nonlinearity affect practical performance.
  • Hardware implementation: Experimental SM implementation confirms applicability in indoor line-of-sight channels and reveals effects from imperfect channel estimation, spatial correlation, and I/Q imbalance.
  • Hardware implementation: Low-resolution-ADC SM detection is reported as robust to quantization error and channel correlation, while hardware studies quantify impairment effects and implement a detector in CMOS.
  • SWIPT: SWIPT-enabled SM uses inactive antennas or relays for energy harvesting through power splitting, simultaneous decoding and harvesting, or distributed protocols.

VI. APPLICATIONS OF SM IN NEW DOMAINS

SM generalizes beyond antenna-based transmission by using activation states or permutations of transmission entities to encode index information across single or multiple signal domains. This produces variants spanning space, frequency, time, code, angle, and combined dimensions.

  • SM generalization: SM conveys additional information through ON/OFF states or permutations of transmission entities rather than only conventional carrier properties.The transmission entity may be physical, such as an antenna, or virtual, such as a space-time matrix.
  • Spatial-domain variants: Spatial-domain variants evolved from single-active-antenna SM and SSK toward schemes supporting multiple active antennas, including (G)SM, (G)SSK, and QSM.These variants were developed to accommodate more than one active antenna and further enhance spectral efficiency.
  • Frequency-domain extensions: The SM concept extends to frequency-domain systems such as SIM-OFDM and MIMO-OFDM-IM, with the latter targeting higher data rates and better error performance than traditional MIMO-OFDM.Low-complexity and near-optimal demodulation algorithms were developed for MIMO-OFDM-IM.
  • Time, code, and angle domains: OFDM-IM helped propagate index modulation to time, code, and angle domains through schemes such as SC-IM and CIM-SS.SC-IM activates a subset of symbols in each time-domain subframe, while CIM-SS targets improved performance with reduced energy consumption.
  • Multi-dimensional designs: Multi-dimensional SM integrates resources across domains to form composite transmission entities, including joint space-time and space-frequency designs.STSK selects among space-time dispersion matrices to compromise between diversity and multiplexing gains, while GSFIM combines space and frequency dimensions.

VII. CONCLUSIONS AND FUTURE DIRECTIONS

The survey presents SM as a promising technology for emerging wireless systems and identifies open problems involving scalability, mobility, channel acquisition, bit-error imbalance, IoT integration, and competing communication requirements. Future work must address these challenges while preserving SM’s energy- and spectral-efficiency benefits.

  • Conclusions: SM offers potential high energy efficiency, low deployment cost, no interchannel interference, relaxed inter-antenna synchronization, and compatibility with massive MIMO.The survey attributes these prospects to conveying information through transmit-antenna activation states with a simple design philosophy.
  • Future directions: Multiuser SM networks require scheduling, resource allocation, antenna allocation, and power allocation to maximize throughput, balance fairness, and manage interference.The survey identifies scalability and integration into multiuser networks as indispensable research tasks.
  • Future directions: High-mobility SM systems in V2X, UAV, and UWA communications require careful pilot-number and pilot-pattern design because pilot occupation consumes space/time/frequency/code resources.Distinguishable pilot patterns may also carry information bits and compensate for pilot overhead.
  • Future directions: Massive-MIMO SM faces channel-acquisition and channel-estimation challenges because training overhead is proportional to the number of antennas, while link-adaptive systems also face difficult channel feedback.The resulting overhead may become prohibitive or unaffordable in practice.
  • Future directions: SM must address unequal spatial-bit and constellation-bit error performance, whose difference depends strongly on the differentiability of channel signatures.Balancing the two error types and scheduling the information bits remain essential problems.
  • Future directions: SM is a promising IoT technique because antenna-state indexing can provide energy-efficient, cost-effective connectivity for massive numbers of constrained, low-rate MTC devices.IoT-specific SM design and integration remain open research directions.
  • Future directions: A fundamental open problem is balancing SM’s energy and spectral efficiency against reliability, capacity, flexible design, and low latency.The survey frames this as a continuing trade-off rather than a resolved design objective.
  • Future directions: Applying or generalizing SM to emerging transmission entities such as orbital angular momentum introduces additional challenging research problems.The survey lists this extension among directions for broadening SM applications.
Loading 1907.02941v1…