Source-linked AI summary
Uplink Contention Based SCMA for 5G Radio Access
Kelvin Au, Liqing Zhang, Hosein Nikopour, Eric Yi, Alireza Bayesteh, Usa Vilaipornsawai, Jianglei Ma, Peiying Zhu
TL;DR
Massive 5G connectivity exposes LTE uplink limitations in signaling overhead, latency, and support for many users. The paper proposes contention-based uplink SCMA with adjustable overloading and evaluates it against OFDMA, finding about a 2.8-times gain in supported active users under low-latency traffic.
Problem
Current LTE systems do not efficiently support massive uplink connectivity, while 5G requires low signaling overhead, low latency, and support for large numbers of users.
Method
The paper proposes contention-based uplink SCMA with adjustable overloading, codebook-based nonorthogonal access, and grant-free transmission enabled by contention regions and blind detection.
Results
Around 2.8 times gain over contention-based OFDMA is demonstrated in supported active users for small-packet, low-latency traffic.
Takeaways & Limitations
Uplink contention-based SCMA can be a promising 5G technology for low-signaling-overhead, low-delay data transmission with massive connectivity.
Takeaways & Limitations
The stated motivation is bounded by LTE's need to reduce resource-allocation and control-signaling overhead, transmission latency, and support for large numbers of users.
Abstract
from arXiv · showhide
Fifth generation (5G) wireless networks are expected to support very diverse applications and terminals. Massive connectivity with a large number of devices is an important requirement for 5G networks. Current LTE system is not able to efficiently support massive connectivity, especially on the uplink (UL). Among the issues arise due to massive connectivity is the cost of signaling overhead and latency. In this paper, an uplink contention-based sparse code multiple access (SCMA) design is proposed as a solution. First, the system design aspects of the proposed multiple-access scheme are described. The SCMA parameters can be adjusted to provide different levels of overloading, thus suitable to meet the diverse traffic connectivity requirements. In addition, the system-level evaluations of a small packet application scenario are provided for contention-based UL SCMA. SCMA is compared to OFDMA in terms of connectivity and drop rate under a tight latency requirement. The simulation results demonstrate that contention-based SCMA can provide around 2.8 times gain over contention-based OFDMA in terms of supported active users. The uplink contention-based SCMA scheme can be a promising technology for 5G wireless networks for data transmission with low signaling overhead, low delay, and support of massive connectivity.
I. INTRODUCTION
The paper motivates contention-based SCMA for 5G uplink massive connectivity because LTE incurs latency and signaling overhead, especially for small or bursty packets.
- 5G requires support for diverse services and massive connectivity among many devices.
- The proposed approach employs contention-based SCMA to address massive-connectivity problems through a new multiple-access design.SCMA is codebook-based, non-orthogonal, and supports overloading with many SCMA layers.
- LTE uplink request-grant scheduling can delay transmission because scheduling requests occur periodically.In the best case described, data may wait 7 ms between the request and uplink transmission.
- Dynamic uplink grants consume downlink control resources, making signaling overhead high relative to small-packet payloads.An uplink grant averaging 2 control channel elements is equivalent to 72 resource elements.
- Semi-persistent scheduling reduces dynamic signaling overhead but is better suited to periodic traffic than bursty traffic.
- Conventional contention-based transmission avoids grant waiting but supports only one simultaneous user in non-MIMO operation.The passage notes that multi-user MIMO increases simultaneous connections, but further work is needed for 5G massive connectivity.
B. UL Contention Based Data Transmission Overview
Contention-based uplink transmission removes the request-grant exchange, while SCMA adds multiplexed, overloaded layers within predefined contention regions.
- Contention-based uplink transmission lets UEs contend for predefined resources without the request-grant procedure.This is intended to reduce uplink latency and signaling overhead for applications with stringent latency requirements.
- Grant-based OFDMA or SC-FDMA orthogonally assigns resources, whereas grant-less operation uses predefined contention regions.
- Contention-based SCMA addresses high-load degradation by incorporating overloading and nonorthogonal transmission.
- SCMA multiplexes multiple user layers within a contention region, with each layer spread across the region’s resources.
- Each SCMA user is associated with a codebook, contention region, and pilot sequence.
III. SYSTEM MODEL
The system maps each user’s bits to a sparse SCMA codeword, multiplexes users over shared resources, and detects the resulting collisions using sparsity-aware processing.
- An SCMA encoder maps log_2(M) bits to a K-dimensional complex codeword selected from a codebook of size M.Each codeword has fewer than K non-zero entries and is transmitted over K radio resources.
- Multiple active users share the same time-frequency resources, so their SCMA layers can collide within an SCMA block.The number of simultaneous layers varies with traffic loading and packet retransmissions.
- The received SCMA-block vector combines users’ codewords, channel effects, received powers, noise, and out-of-cell interference.
- SCMA sparsity reduces the number of non-zero colliding elements per tone, enabling MAP multi-user detection with limited complexity.With two receive antennas, detecting six SCMA layers at 150% overloading requires less than four times the complexity of a linear MMSE receiver.
B. SCMA Scalability for Massive Connectivity
SCMA scalability comes from generating multiple sparse codebooks over K resources and adjusting codeword sparsity to control overloading.
- The maximum number of codebooks J depends on the number of non-zero entries N and codeword length K.Selecting N non-zero positions within K elements is a combination problem.
- The overloading factor is defined by multiplexing J codewords over K resources.
- Adjusting K and N enables different overloading levels and different numbers of codebooks.The paper states that massive connectivity favors an overloading factor much greater than 1.
- For K=4 and N=2, six codebooks can be generated; two input bits are mapped to a codeword and spread over four subcarriers.Multiple users’ data streams are then overlaid using codewords from different codebooks.
- Fig. 1 illustrates SCMA codebooks, encoding, and multiplexing.
IV. PHY AND MAC LAYER DESIGN
The proposed SCMA resource structure defines contention transmission units by combining time-frequency resources with codebooks and pilots. Codebook reuse allows more active users, while collisions occur when users select the same pilot sequence.
- UL SCMA Resource Definition: A contention transmission unit combines time, frequency, an SCMA codebook, and a pilot sequence.For J codebooks and L pilots per codebook, the time-frequency region contains L×J CTUs.
- UL SCMA Resource Definition: Multiple terminals can share a codebook, and distinct pilot sequences enable the receiver to separate their data streams.The MPA receiver estimates users’ channels through their different pilots.
- UL SCMA Resource Definition: User collision occurs when two or more users select the same pilot sequence within a contention region.Such pilot collisions are resolved using random back-off.
- UL SCMA Resource Definition: Contention regions consist of time-frequency resources over which SCMA codebooks are overlaid.Their size and number depend on factors including the expected number of terminals and suitable applications.
B. Transmission Mechanism
Contention-based SCMA lets multiple UEs transmit using predefined resources and codebooks without dynamic uplink grants. It uses blind detection, timing procedures, and random back-off to manage access and collisions.
- Transmission Mechanism: The scheme combines contention-based multiple access, where multiple UEs may share resources, with predefined non-adaptive coding and modulation levels.This removes the need to assign each transmission a unique scheduled resource.
- Transmission Mechanism: UEs transmit in predefined contention regions using CTUs assigned explicitly by signaling or implicitly from UE identities.A UE-to-CTU mapping can use the UE ID modulo the total number of resources and vary over time for collision diversity.
- Transmission Mechanism: The network attempts reception with all access codes in the contention region and may use joint data-activity detection, CRCs, or headers.These mechanisms also support detection decisions and HARQ retransmissions.
- Transmission Mechanism: Uplink timing uses an LTE-like timing-advance procedure before transmission when timing alignment has been lost.In mixed traffic, scheduled transmissions can maintain uplink timing alignment.
- Transmission Mechanism: Colliding transmissions are retransmitted after each UE selects a random back-off time from a back-off window.Retransmission uses the predefined CTU as the original transmission.
- Transmission Mechanism: Predefined regions, CTU assignments, and blind detection eliminate dynamic uplink grants for contention-based SCMA.Because uplink grants occupy downlink control resources, this can reduce downlink signaling overhead.
V. NUMERICAL RESULTS AND ANALYSIS
The paper evaluates SCMA’s scalability and its potential gain over contention-based OFDMA through system-level uplink simulations.
- Numerical Results and Analysis: System-level simulations assess SCMA scalability and its potential uplink gain over contention-based OFDMA.The evaluation targets a small-packet application scenario.
A. Scalability Analysis
SCMA scalability is analyzed through the number of codebooks generated from codeword sparsity and length, followed by a small-packet system-level simulation under low-latency conditions.
- Scalability Analysis: The maximum number of SCMA codebooks is determined by the number of non-zero codeword entries, N, and codeword length, K.The possible non-zero positions are counted using a binomial coefficient.
- Scalability Analysis: 70 codebooks can be generated when K=8 and N=4.Varying the spreading factor and the number of non-zero entries can increase the codebook count dramatically.
- Simulation Setup and Assumptions: The simulation uses a 19-cell, 3-sector network at a 2GHz carrier frequency.Users are dropped randomly, and statistics are collected for a sector.
- Simulation Setup and Assumptions: Each transmission carries the same data size fitting one OFDMA resource-block pair, with fixed spectral efficiency of 1 bit/s/Hz.SCMA spreads data over four RB pairs, whereas OFDMA uses one of the four RB pairs.
- Simulation Setup and Assumptions: Packets are dropped when their waiting and transmission time exceeds the latency requirement.The simulation assumes perfect channel estimation.
C. Performance and analysis
Under a 5 ms latency requirement without retransmissions, contention-based SCMA maintains lower packet-drop rates and supports more active users than OFDMA at comparable outage targets.
- A 5 ms delay bound is evaluated without any retransmission opportunity for contention-based UL SCMA and OFDMA.The evaluation targets small-packet transmission under tight latency requirements.
- SCMA has lower user packet-drop-rate distributions than OFDMA across different traffic loadings.OFDMA packet-drop performance degrades faster as traffic load increases.
- At 2% system outage, SCMA supports 119 users compared with 42 users for OFDMA.The comparison uses the number of supported active users at the same system outage target.
- SCMA gains 2.83 and 2.74 times over OFDMA in supported users at system outages of 2% and 5%, respectively.These are simulated outage-capacity gains based on supported-user counts.
D. PAPR Reduction for SCMA
SCMA-OFDM inherits the PAPR concern associated with OFDMA, but codebook design provides an additional degree of freedom for reducing it.
- SCMA-OFDM can address high PAPR during SCMA codebook design, unlike original OFDM.Low-projection codebooks guarantee PAPR below 5 dB in certain narrowband scenarios.
- Designed low-projection codebooks achieve PAPR below 5 dB for SCMA-OFDM in certain narrowband scenarios.This is comparable with the current low-PAPR SC-FDMA waveform.
VI. CONCLUSION
The paper proposes uplink contention-based SCMA with adjustable overloading and evaluates it for small-packet, low-latency transmission. Simulations report roughly 2.8 times the supported-user gain over contention-based OFDMA.
- The paper proposes an uplink contention-based SCMA scheme for 5G wireless networks.The system design aspects of the new multiple-access scheme are described.
- SCMA parameters can be adjusted to provide different overloading levels for diverse 5G traffic-connectivity requirements.
- System-level evaluations compare contention-based UL SCMA with OFDMA for small-packet transmission under tight latency requirements.
- Around 2.8 times gain is reported for SCMA over OFDMA in supported active users in a low-latency contention-based system.
- The proposed scheme is identified as a promising 5G technology for low-signaling-overhead, low-delay data transmission with massive connectivity.