Source-linked AI summary
Harvest-Then-Cooperate: Wireless-Powered Cooperative Communications
He Chen, Yonghui Li, Joao Luiz Rebelatto, Bartolomeu F. Uchoa-Filhoand, Branka Vucetic
TL;DR
The paper studies cooperative communication when the source and relay lack embedded energy supplies and must harvest power from an AP. It proposes HTC, derives approximate throughput expressions for single- and multi-relay Rayleigh-fading networks, and validates the analyses numerically, finding improved performance over harvest-then-transmit in simulated scenarios.
Problem
The paper addresses cooperative communication in networks whose source and relay have no embedded energy supply and must rely on wireless energy harvested from an AP.
Method
The paper proposes HTC, analyzes a three-node model and its multi-relay extension with opportunistic relaying and partial relay selection, and derives approximate throughput expressions over Rayleigh fading.
Results
The proposed HTC protocol outperforms harvest-then-transmit in all simulated scenarios, with performance further improving as the number of relays increases.
Takeaways & Limitations
Throughput depends on time allocation, relay number, and relay position; opportunistic relaying achieves the best throughput with optimal time allocation and relay position.
Abstract
from arXiv · showhide
In this paper, we consider a wireless-powered cooperative communication network consisting of one hybrid access-point (AP), one source, and one relay. In contrast to conventional cooperative networks, the source and relay in the considered network have no embedded energy supply. They need to rely on the energy harvested from the signals broadcasted by the AP for their cooperative information transmission. Based on this three-node reference model, we propose a harvest-then-cooperate (HTC) protocol, in which the source and relay harvest energy from the AP in the downlink and work cooperatively in the uplink for the source's information transmission. Considering a delay-limited transmission mode, the approximate closed-form expression for the average throughput of the proposed protocol is derived over Rayleigh fading channels. Subsequently, this analysis is extended to the multi-relay scenario, where the approximate throughput of the HTC protocol with two popular relay selection schemes is derived. The asymptotic analyses for the throughput performance of the considered schemes at high signal-to-noise radio are also provided. All theoretical results are validated by numerical simulations. The impacts of the system parameters, such as time allocation, relay number, and relay position, on the throughput performance are extensively investigated.
I. INTRODUCTION
The paper introduces a wireless-powered cooperative network in which source and relay nodes harvest downlink energy from an AP before cooperatively transmitting uplink information. It develops HTC throughput analyses for single- and multi-relay settings and validates their performance numerically.
- System motivation and model: Wireless-powered cooperative communication uses downlink energy transfer from a hybrid AP and cooperative uplink information transmission from energy-harvesting source and relay nodes.The source and relay have no other embedded energy supply and store harvested energy in rechargeable batteries.
- Proposed protocol: The HTC protocol assigns the source and relay to harvest energy during the downlink and cooperate during the uplink, using AF relaying and selection combining.The protocol is analyzed in delay-limited transmission mode over Rayleigh fading channels.
- Multi-relay extension: The analysis extends from the three-node reference model to multi-relay networks using opportunistic relaying and partial relay selection.Only one relay is selected in each transmission block according to a specified criterion and forwards the source signal using energy harvested in that block.
- Performance evaluation: Numerical simulations validate the theoretical results and investigate how time allocation, relay number, and relay position affect throughput.The proposed scheme outperforms the existing harvest-then-transmit protocol in all simulated cases.
- Analytical distinction: Energy transfer makes uplink source transmit power random and induces mutual correlation among source-AP and source-relay-AP link SNRs.Consequently, analytical tools developed for conventional cooperative networks cannot be directly applied.
II. SYSTEM MODEL AND PROTOCOL DESCRIPTION
The system uses downlink wireless energy transfer followed by cooperative uplink information transmission. The HTC protocol allocates time between harvesting and two equal-length uplink slots, with source and relay powers determined by harvested energy.
- System model: The network has a single-antenna AP, source, and relay; source and relay lack other energy supplies and store harvested AP energy in rechargeable batteries.All nodes operate in half-duplex mode.
- Channel and energy assumptions: The model assumes independent slow, frequency-flat fading with channel gains constant within each transmission block.Noise harvesting is neglected under sufficiently large AP transmit power.
- HTC protocol: Each block assigns τT to AP-to-terminal energy transfer and divides the remaining (1−τ)T equally between two cooperative uplink slots.The source transmits first, and the relay forwards using amplify-and-forward in the second slot.
- HTC protocol: The source and relay exhaust their harvested energy for uplink transmission, although cross-block power allocation could alternatively be used.The paper assumes current-block energy exhaustion for analysis.
- Power allocation: The source and relay transmit with powers proportional to ητPA and their AP-link channel gains, scaled by 2/(1−τ).These powers follow directly from harvested energy divided by each uplink slot duration.
- Cooperative reception: The relay amplifies the source’s overheard signal and forwards it to the AP, while selection combining combines the direct and relayed receptions.The relay uses its harvested power and an amplify-and-forward factor.
III. THROUGHPUT ANALYSIS OF THE HTC PROTOCOL
The paper analyzes HTC throughput through fixed-rate outage probability in delay-limited transmission and derives approximate closed-form expressions. It also compares HTC with harvest-then-transmit and examines the time-allocation trade-off.
- Throughput formulation: For transmission rate R, outage probability Pout, and transmission duration t, throughput is R(1−Pout)t.This fixed-rate formulation is used for delay-limited transmission.
- HTC analysis: The analysis first characterizes HTC outage probability and then derives an approximate closed-form average-throughput expression for the three-node model.The outage expression is stated in Proposition 1 and the throughput expression in Corollary 1.
- Baseline comparison: The harvest-then-transmit baseline is analyzed separately without relay assistance, yielding its own average-throughput expression.The baseline uses the source’s direct uplink transmission.
- Time allocation: Increasing τ lowers outage probability because it increases harvested energy, but excessive τ reduces uplink transmission time and can lower throughput.Consequently, the throughput is expected to have an interior optimal τ between 0 and 1.
- Cooperative benefit: HTC can improve performance over harvest-then-transmit when the source-AP link is poor and the relay can help deliver the source information.If the AP-source link is poor enough to reduce harvested source energy, both source-AP and source-relay links may experience outage.
IV. EXTENSION TO MULTI-RELAY SCENARIO
The paper extends HTC analysis to multiple relays and considers opportunistic relaying and partial relay selection. Relay selection uses either both hops jointly or channel-state information from one hop.
- Multi-relay model: The multi-relay extension considers N relays positioned between the source and AP, with location-based clustering yielding equivalent average channel power gains.The relays are assumed to be relatively close together.
- Motivation: OR and PRS are analyzed as representative AF relay-selection schemes that improve network performance while avoiding reduced spectral efficiency in conventional multi-relay cooperation.The paper extends the HTC throughput analysis to these two selection strategies.
- OR protocol: Opportunistic relaying selects the relay maximizing min(γSRi, γRiA), jointly accounting for the source-relay and relay-AP hops.The selected relay is denoted Rb.
- PRS protocol: Partial relay selection uses channel-state information from only one hop, selecting a relay based on either the source-relay or relay-AP link.The first-hop and second-hop variants are described separately.
- Analysis assumptions: The selected relay exhausts the energy harvested during the current block to forward the source’s signal.The paper derives approximate and high-SNR asymptotic throughput expressions for both OR and PRS.
A. OR Protocol
The OR analysis derives an approximate outage probability for multi-relay HTC. This result supports subsequent throughput analysis for opportunistic relay selection.
- OR outage analysis: The paper first derives the outage probability of HTC with OR before analyzing its throughput.The derivation is presented as Proposition 2.
P HTC,OR
The HTC protocol’s OR analysis derives approximate throughput expressions and examines their high-SNR behavior. The three-node reference result follows by setting N = 1.
- The derivation is validated because (17) simplifies to (8) when N = 1.
- The section gives an approximate throughput expression for the HTC protocol with OR.
- The analysis investigates asymptotic throughput with OR at sufficiently high PA/N0.
- The asymptotic throughput of the three-node reference model is obtained by substituting N = 1 into (19).
B. PRS protocol
The PRS analysis derives approximate outage and throughput expressions for two relay-selection criteria, including their asymptotic performance. The schemes select one relay per block, while energy accumulation is identified as future work that could improve performance.
- PRS protocol: Approximate outage-probability expressions are derived for the two PRS protocols defined by the selection criteria in (15) and (16).
- PRS protocol: The considered system’s approximate PRS throughput is given for X ∈ {I, II}, representing the two selection schemes.
- PRS protocol: The PRS throughput expression is also analyzed asymptotically.
- PRS protocol: Only one relay is selected per block; non-selected relays can store harvested energy for future transmission.
- PRS protocol: Considering accumulated relay energy could further improve performance but requires a more sophisticated scheduling scheme, left for future work.
V. NUMERICAL RESULTS
Numerical results validate the throughput analysis and examine time allocation, relay number, and relay position. The HTC protocol outperforms harvest-then-transmit in all simulated cases, with relay selection and placement affecting performance.
- Simulation setup: The simulations use a linear relay topology and fixed channel parameters, including dAS = 10m, χ = 2, N0 = −80dBm, η = 0.5, and R = 1 bpcu.Relay distance determines both received-signal reliability and energy attenuation in the adopted channel model.
- Validation: The approximate throughput expression becomes very tight in medium and high SNR conditions, agreeing well with simulation results.The asymptotic expression also approaches the simulations in the reported high-power range.
- Validation: At sufficiently high SNR, HTC throughput saturates at the maximal value R (1 − τ) as outage probability approaches zero.This saturation applies for a fixed time-allocation parameter τ.
- Time allocation: Each scheme has an optimal τ in the two-relay case, obtained by one-dimensional exhaustive search for PA = 35dBm and dSR = 3m.The throughput is plotted against τ to study this allocation trade-off.
- Time allocation: Optimal τ decreases monotonically as PA increases; opportunistic relaying has the smallest optimum, while harvest-then-transmit has the largest.The two partial-relay-selection schemes fall between them and can exchange ordering with relay position.
- Relay number: With optimal τ, the three relay-selection schemes coincide at N = 1, while increasing relay number improves HTC performance with diminishing gains, especially for partial relay selection.Harvest-then-transmit is included as the N = 0 special case.
- Relay position: Opportunistic relaying always performs best in the reported relay-number comparisons, whereas PRS-II exceeds PRS-I near the source and the ordering reverses when relays are farther away.The selection criterion should emphasize the second hop when relays are close to the source and the first hop when they are sufficiently far.
VI. CONCLUSIONS
The paper proposes the HTC protocol for wireless-powered cooperative communication and derives approximate throughput expressions for single- and multi-relay settings. Numerical results show HTC advantages across simulated scenarios and identify relay-selection and parameter conditions associated with stronger throughput.
- The HTC protocol is proposed for wireless-powered cooperative communication networks, with an approximate closed-form average-throughput expression derived over Rayleigh fading channels.
- The analysis extends to multiple relays using opportunistic relaying and partial relay selection schemes.
- Numerical results show that HTC outperforms harvest-then-transmit in all simulated scenarios.
- Increasing the number of relays can further improve HTC's performance.
- With optimal time allocation and relay position, opportunistic relaying achieves the best throughput, while second-hop partial relay selection can outperform first-hop selection.
A. Proof of Proposition 1
The proof formulates HTC outage through the direct and relayed-path SNRs, whose shared AP-to-source channel creates correlation. An analytically tractable SNR approximation enables the subsequent closed-form outage derivation.
- The source-to-AP mutual information is used to formulate the outage probability through the threshold event γA < ν.
- The direct and relayed SNRs are correlated because both contain the random variable hAS, unlike the normally independent path SNRs in constant-powered cooperative networks.
- An approximation replaces the complex relayed-path SNR with a more tractable expression for deriving the outage probability.The text states that the approximation is accurate enough at medium and high SNR values.
- The approximate outage expression is evaluated by calculating three component probabilities and substituting them into the outage relation.
B. Proof of Proposition 2
The proof derives the opportunistic-relaying outage expression using the relay-selection criterion, independence properties, and integral expansions. Substitution and integral evaluation yield the stated throughput result.
- The opportunistic-relaying outage probability is approximated using its selection criterion and the SNR approximation.
- Independence and identical-distribution properties of the source-relay, AP-relay, and relay-AP channel sets simplify the term T2.
- The term T3 is combined with a binomial expansion of T2 to obtain a further outage-expression form.
- Substituting the derived terms and solving the resulting integrals produces the desired opportunistic-relaying result.
C. Proof of Proposition 3
The proof develops the high-SNR approximation by analyzing S(x) as x approaches zero, then applies it to the opportunistic-relaying throughput. Related partial-relay-selection expressions are obtained through analogous outage and integral calculations.
- High-SNR asymptotics: At high SNR, the terms inside S(·) approach zero, so its behavior is analyzed for x → 0.
- High-SNR asymptotics: The approximation of S(x) is derived using a series representation of the modified Bessel function and the small-argument behavior of I1(x).
- High-SNR asymptotics: Exact and asymptotic values of S(x) coincide closely as x approaches zero, validating the approximation in (42).
- High-SNR asymptotics: Using the S(x) approximation and simplifications yields the high-SNR throughput expression for opportunistic relaying.
- Partial relay selection: For first-hop partial relay selection, the selected relay distribution is derived from the source-relay selection criterion, while AP-relay and relay-AP links remain unselected.
- Partial relay selection: For second-hop partial relay selection, the selected-relay condition changes and the resulting expression follows by repeating the outage-integral calculation.