Source-linked AI summary
NOMA based Random Access with Multichannel ALOHA
Jinho Choi
TL;DR
NOMA is usually coordinated and downlink-focused, creating a gap for uncoordinated uplink random access. The paper applies predetermined-power-level NOMA to multichannel ALOHA and reports higher throughput without bandwidth expansion, especially when subchannels are limited. It also studies channel-dependent selection to reduce transmission power.
Problem
NOMA generally relies on coordinated power allocation and full CSI, limiting its apparent suitability for uncoordinated random access.
Method
The paper applies a predetermined-power-level NOMA scheme to multichannel ALOHA and studies channel-dependent subchannel and power-level selection.
Results
With L = 4, throughput is 3 times higher than conventional multichannel ALOHA at fixed B, while larger L eventually yields limited improvement.
Takeaways & Limitations
NOMA can effectively increase available subchannels and improve random-access throughput without bandwidth expansion when subchannels are limited.
Takeaways & Limitations
The scheme’s main drawback is high transmission power, addressed through channel-dependent selection and an average-power upper bound.
Abstract
from arXiv · showhide
In nonorthogonal multiple access (NOMA), the power difference of multiple signals is exploited for multiple access and successive interference cancellation (SIC) is employed at a receiver to mitigate co-channel interference. Thus, NOMA is usually employed for coordinated transmissions and mostly applied to downlink transmissions where a base station (BS) per- forms coordination for downlink transmissions with full channel state information (CSI). In this paper, however, we show that NOMA can also be employed for non-coordinated transmissions such as random access for uplink transmissions. We apply a NOMA scheme to multichannel ALOHA and show that the throughput can be improved. In particular, the resulting scheme is suitable for random access when the number of subchannels is limited since NOMA can effectively increase the number of subchannels without any bandwidth expansion.
I. INTRODUCTION
NOMA is typically coordinated and downlink-oriented because it relies on power differences, SIC, and CSI, but this paper applies it to uncoordinated uplink random access. The proposed combination with multichannel ALOHA improves throughput without bandwidth expansion and can help when subchannels are limited.
- NOMA shares radio resources among users by exploiting transmission-power differences and SIC to mitigate co-channel interference.
- Random access is suitable for short-packet MTC uplinks because it reduces signaling overhead.
- Known CSI and coordinated power allocation make NOMA easier for downlink than uncoordinated uplink transmissions.
- The paper applies predetermined-power-level NOMA to multichannel ALOHA for uncoordinated uplink random access.
- The proposed scheme improves throughput without bandwidth expansion and effectively increases available subchannels when bandwidth is limited.
- A channel-dependent subchannel and power-level selection scheme is studied to reduce the proposed scheme’s potentially high transmission power.
II. MULTICHANNEL ALOHA
Multichannel ALOHA extends slotted ALOHA with orthogonal subchannels for uplink random access. Under a collision-model analysis, its average throughput increases with the number of subchannels and reaches Be^-1 at optimal intensity.
- The analysis assumes one base station, multiple users, and a single-cell uplink setting.
- Multichannel ALOHA generalizes ALOHA by providing multiple orthogonal subchannels for uplink transmissions.
- Each active user independently and uniformly selects a subchannel, while throughput analysis ignores capture and uses a simple collision model.
- With K users active independently at probability p_a, the number selecting a subchannel is approximated by a Poisson random variable for large K.
- Be^-1 is the maximum average throughput at intensity λ = 1.
- Throughput is defined as the average number of users successfully accessing a channel without collision.
III. RANDOM ACCESS BASED ON NOMA
The paper presents a single-subchannel random-access scheme based on a previously studied NOMA approach and derives its conditional throughput.
- The section considers one subchannel, omits the subchannel index, and derives conditional throughput for a NOMA-based random-access scheme.
A. A NOMA Scheme: Power Division Multiple Access
The proposed PDMA-based NOMA scheme uses predetermined power levels so simultaneous signals can be separated by SIC. It can decode multiple signals, but large power-level counts increase power demands and error-propagation risk.
- A. A NOMA Scheme: Power Division Multiple Access: The scheme differs from conventional coordinated uplink NOMA by avoiding base-station power allocation with full CSI.
- A. A NOMA Scheme: Power Division Multiple Access: Each active user randomly selects one of L ordered power levels, with transmission power adapted to its channel gain.
- A. A NOMA Scheme: Power Division Multiple Access: The power levels are designed so that, when every level is occupied, SIC can decode signals in ascending power order at target SINR Γ.
- A. A NOMA Scheme: Power Division Multiple Access: L signals can be decoded despite simultaneous transmission when the rate is R = log2(1 + Γ).
- A. A NOMA Scheme: Power Division Multiple Access: If M ≤ L active users choose distinct power levels, all M signals can be decoded; this is called power-domain multiple access (PDMA).
- A. A NOMA Scheme: Power Division Multiple Access: Ideal SIC and capacity-achieving codes are assumed, while practical decoding errors can cause error propagation and make large L undesirable.
B. Throughput Analysis
PDMA can decode all active-user signals when users select distinct power levels, but power collisions create dependent decoding failures. Increasing the number of power levels raises throughput while sharply increasing transmission power, limiting practical gains.
- PDMA successfully decodes all M active-user signals when M ≤ L and all users choose different power levels.When multiple users select the same power level, a power collision occurs and higher-level signals may become undecodable.
- Power collisions are dependent: a collision at level l prevents decoding signals at levels l + 1 through L.
- If M = 2, the lower-bound on conditional throughput is exact because either both signals use different power levels or neither is decoded.
- Higher L increases PDMA throughput, but the highest power level grows exponentially and average transmission power rises significantly.The throughput increase is not significant as L increases, making large L potentially impractical.
IV. APPLICATION OF NOMA TO MULTICHANNEL ALOHA
The paper applies PDMA to multichannel ALOHA as NM-ALOHA and also studies channel-dependent selection of subchannels and power levels. The selection scheme targets lower transmission power or improved energy efficiency.
- NM-ALOHA applies PDMA to multichannel ALOHA and studies its throughput.
- Channel-dependent selection of subchannels and power levels is studied to reduce transmission power or improve energy efficiency.
A. Application of PDMA to Multichannel ALOHA and Throughput Analysis
NM-ALOHA assigns PDMA across multichannel ALOHA subchannels and analyzes its conditional and average throughput. With L = 2, it achieves 1.5 times the throughput of standard multichannel ALOHA, while larger L increases transmission power.
- PDMA is applied to every one of B ALOHA subchannels, creating LB effective subchannels.
- The conditional throughput of NM-ALOHA is defined for a given active-user allocation across the B subchannels.
- Uniform random subchannel selection yields a conditional-throughput expression whose subchannel occupancy variable N is binomial with p = 1/B.
- For large K, Poisson approximation provides a lower bound on NM-ALOHA’s average throughput.
- With L = 2, the lower bound is exact, and NM-ALOHA throughput is 1.5 times higher than standard multichannel ALOHA.
- Because the conditional-throughput term increases with L, NM-ALOHA’s average-throughput lower bound also increases with L without bandwidth expansion.
- Increasing L also increases transmission power, motivating channel-dependent subchannel and power-level selection.
B. Channel-Dependent Energy Efficient Selection
The channel-dependent selection scheme uses channel conditions to choose subchannels and power levels, reducing transmission power without affecting throughput. Its energy-efficiency advantage over random selection grows with the number of power levels, while B > 1 mitigates the unbounded-power issue.
- Selection scheme: Channel-dependent selection chooses subchannels and power levels using channel gains, whereas the baseline selects them independently and uniformly at random.The selection can use the maximum channel gain among subchannels to further reduce transmission power.
- Selection scheme: Users are divided into groups according to large-scale fading or distance, with each group assigned a predetermined power level.For L = 2, users farther from the BS tend to choose a smaller power level to reduce overall transmission power.
- Power boundary: Average transmission power can become unbounded as B approaches 1, whereas for B ≥ 2 the corresponding expectation is finite.The paper identifies B = 1 as the case producing prohibitively high transmission power and discusses truncated channel inversion as a mitigation.
- Energy-efficiency effect: The channel-dependent selection scheme does not affect throughput when user locations are random, but it can greatly improve energy efficiency.Its average transmission power grows more slowly with L than under random subchannel and power-level selection for B > 1 and L > 1.
V. SIMULATION RESULTS
Simulations show that NM-ALOHA improves throughput by exploiting multiple power levels without bandwidth expansion, especially when the number of subchannels is limited. Channel-dependent selection reduces transmission power, while throughput has an optimal access probability and the lower-bound is reasonably tight.
- Throughput versus subchannels: 3.5 throughput is achieved by NM-ALOHA with L = 4 and B = 4, matching multichannel ALOHA with B = 10.The paper reports this as 2.5 times higher spectral efficiency for NM-ALOHA than conventional multichannel ALOHA.
- Throughput versus subchannels: About 4 times higher throughput is achieved by NM-ALOHA than multichannel ALOHA at B = 4, with the gap decreasing as B increases.Throughput increases with the number of subchannels for both compared schemes.
- Throughput versus power levels: 3 times higher throughput is obtained with L = 4 than with conventional multichannel ALOHA at fixed B = 6.Throughput increases with L without bandwidth expansion, although the improvement becomes limited when L is sufficiently large.
- Throughput versus access probability: Throughput first increases and then decreases with access probability p_a, implying an optimal access probability that maximizes throughput.The paper notes that access control using p_a or the number of subchannels is beyond its scope.
- Bounds: The lower-bound from (14) is reasonably tight and becomes tighter as L decreases.This conclusion is drawn from Figs. 2–4.
- Transmission power: Average transmission power increases with L, while channel-dependent selection produces much lower power than channel-independent random selection.The simulations impose a transmission-power limit of 10L dB because channel inversion can otherwise produce arbitrarily high power.
- Transmission power: Average transmission power decreases with B under channel-dependent selection, while random selection does not depend on B.Thus, increasing B can improve energy efficiency even when it does not significantly improve throughput at fixed p_a.
VI. CONCLUDING REMARKS
The proposed NOMA-based multichannel ALOHA scheme increases throughput by using multiple power levels alongside subchannels, making it suitable when subchannels are limited without bandwidth expansion. Its main drawback is high transmission power, which channel-dependent selection can reduce or convert into improved energy efficiency.
- The proposed scheme uses multiple subchannels and power levels to increase the effective number of subchannels without bandwidth expansion.
- It provides higher throughput than multichannel ALOHA by exploiting power differences.
- A closed-form lower-bound expression was derived to evaluate the throughput performance.
- The main drawback is high transmission power associated with exploiting the power domain in NOMA.
- Channel-dependent selection of the subchannel and power level reduces transmission power or improves energy efficiency.