Energy harvesting (EH) has attracted significant interests from academia and industry in recent years, as both the concept of green communication and practical needs from low power sensor networks are growing steadily. There are successful industrial EH products that harvest energy through light, kinetic motion, pressure, etc., for sensors and controls . Harvesting natural energy, e.g., solar, piezoelectric, kinetic and wind energy , in general presents a reliability challenge because of the instability of the natural environment. There is much work in the literature on natural energy harvesting. In , two energy management strategies are proposed – one is to maximize the total throughput within a given time slot, and the other is to minimize the transmission completion time given target data throughputs. In , optimal online and offline energy management policies are designed by applying dynamic programming and staircase water-filling algorithm, respectively. Relative to natural resources of which EH can be greatly affected by environment changes, RF signals are more stable and human-controllable . With rapid development and deployment of various wireless networks, contemporary radio wave energy harvesting techniques have emerged as promising solutions in the dealings with crucial energy constraint in battery-run networks. There are mainly two categories of RF wave based energy supply techniques. The first category is communication systems with simultaneous wireless information and power transfer (SWIPT) first proposed in . Because information and energy receivers are sensitive to different power values, two co-located receiver schemes for SWIPT, known as time switching (TS) and power splitting (PS), are described in . Based on TS and PS receiver schemes, a general PS scheme termed dynamic PS (DPS) is proposed in [7, 8]. The second category is the RF-energy based wireless powered communication networking (WPCN) where wireless devices are charged by RF signals from a hybrid access point (H-AP) with power supply, first proposed in . A “harvest-then-transmit” protocol is introduced in , where the sum-throughput maximization and common-throughput maximization are investigated. Recently, WPCN has emerged as an appealing candidate solution in future self-sustainable wireless communication systems [10, 11].
Furthermore, efforts have been made to design cooperative and efficient energy management strategies based on a three-node network including a relay, where both the EH source and EH relay harvest energy from nature in [12, 13, 14, 15]. In , an energy cooperation strategy is introduced that allows the EH source node to share some energy with the EH relay node. A wireless cooperative transmission scheme with energy salvage is proposed in , where the source can harvest energy from the relay signal transmitted to the destination. Under the decode-and-forward (DF) relaying scheme in , two optimal power allocation algorithms are developed to solve the throughput maximization problems with delay-constrained and no-delay-constrained traffic at the destination, respectively. Furthermore, short-term sum-rate maximization problems are solved in , where a two-way half-duplex relay channel is considered with DF relaying scheme.
Based on SWIPT techniques, the TS and PS protocols are also extended to three-node communication networks with relay assistance in [16, 17, 18]. Considering amplify-and-forwarding (AF) relaying in , the EH relay node has no other energy sources but uses TS or PS to split the received RF signal from the source into two streams, one for EH and the other for information forwarding. Considering both AF and DF schemes, the performance of EH and throughput are analyzed in , where the relay node uses TS to split received RF signals. Furthermore, under the DF relaying strategy, multiple source-destination pairs with only one relay are considered in , where the EH relay adopts PS to harvest energy from the RF signals transmitted from multiple sources. On the other hand, relay assistance is also applied to improve the performance of WPCN. In , “harvest-then-transmit” is extended to “harvest-then-cooperate” scheme that enables a relay node to forward the information transmission of the source node. Different from  where the relay node only transmits the information of the source, in  the “relay” node (the near user to H-AP) transmits its own data and forwards the information of the “far” user, which is called “user cooperation”.
Energy harvesting becomes even more relevant in a wireless sensor network for the rapidly increasing Internet of Things (IoT) applications. Many of these low power sensor nodes are distributed in areas with no power supply and expected to last for years running on battery and/or on EH from all available sources. Because nodes are located in different places, the energy they can harvest may differ as well. For example, a node that resides at a windy spot will harvest more wind energy than a neighboring node surrounded by structures. Therefore, not only data cooperation such as relaying will help to improve the overall performance of the network, RF energy cooperation/sharing will be an important approach to enhance the network operation. Moreover, due to the broadcast nature of wireless, the RF energy from one transmitting node can be harvested by all nodes nearby, regardless if the information is intended for the nodes. It is intuitive that the optimal energy and data cooperation strategy depends on the EH conditions of the nodes and the propagation channels among the nodes. Another consideration is that the order in which the individual nodes transmit their messages with continuous EH supplies appear to impact the system’s performance, because the nodes who transmit later may store more energy, with the potential to perform better than the scenario when they transmit earlier. In this paper, we are motivated to investigate the energy and data cooperation strategies in a multi-node sensor network, where each node harvests energy from natural environments as well as from each other. To enable tractable initial study, we consider a simple three node topology to illustrate the main idea, with potential extensions to more sophisticated network architectures in the future. The main contributions of the paper are summarized below.
A new cooperative wireless communication network strategy is proposed. The model establishment, design goal formulations, and algorithms for throughput maximization under the proposal are carried out and illustrated using a three-node network with two EH user nodes and a destination node. Transmission models are established based on performance analysis for a total of four scenarios, each with two cases to distinguish which of the two users transmits first, that take into account node activity status in terms of whether it transmits (we then call it active) while harvesting surrounding natural energy, or it does not transmit (then call it inactive) while harvesting natural energy as well as RF signals broadcast by the active user. In addition, the power-splitting strategy adopted by the protocol allows a near-to-destination user to properly split the power of the RF signals from the “far” user, with one part for EH and the rest for relaying message to the destination. Collectively, this is a suite of transmission models for wireless networks with EH capabilities, where optimal throughputs can be achieved with the appropriate network cooperation strategy adaptive to the energy harvesting and propagation environments.
Based on the proposed models, the problems of jointly optimizing the distribution of time intervals for data cooperation and energy harvesting, power allocations over these time intervals, and data throughputs at user nodes so as to maximize the sum-throughput or, alternatively, the minimum throughput of the two users are formulated as convex constrained problems in all scenarios.
Two fast algorithms are proposed to solve the aforementioned optimization problems. To be specific, an accelerated Newton barrier (NB) algorithm is developed to solve the constrained optimization problem, and the acceleration is achieved by a tailored line search technique which is required in each NB iteration. An alternative and often more efficient algorithm is developed based on a local approximation of the logarithmic terms involved. The results are convex quadratically constrained quadratic programming (QCQP) problems for several complex scenarios and quadratic programming (QP) problems for the rest of scenarios. Primal-dual path-following interior-point algorithms with closed-form line steps are deduced for these QCQP and QP problems and the efficiency of their implementations is evaluated and compared favorably with several available computer codes.
The rest of the paper is organized as follows. Section II presents the system model that consists of channel model, EH model, and transmission models for four possible scenarios. Section III presents problem formulations based on the four scenarios, with the objective of maximizing the weighed sum-throughput and minimum throughput of the two users, respectively. In Section IV, we describe two convex optimization algorithms to solve the problems formulated in Section III. Numerical results are given in Section V to evaluate how a wireless network in question performs with respect to the variations of the system parameters. Observations made from the simulation results are found supportive to our analysis. Finally, Section VI concludes the paper. We conclude this section with a remark on notation: In what follows, denotes the base-2 logarithm, and denotes the statistical expectation.
Ii System Model
As illustrated in Fig. 1, we consider a three-node network with two EH sensor nodes (called users) and a destination node , each of which is equipped with a single antenna. Let be the distance from to . Without loss of generality, we assume , and name and the near user and far user, respectively. Concerning the power supply status in the network, node , simulating a gateway or data collector in a wireless sensor network, is powered by a stable energy supply, while the two sensor nodes are powered by natural energy resources (e.g., solar, wind) as well as RF signals. The near user has data to send to the destination D, while the far user sends its data either directly to the destination or with the help of acting as a relay. The near user in our model acts as a decode-and-forward (DF) relay node and this message relaying process is called data cooperation (DC). Without data cooperation, must send with higher transmission powert in order to reach , compared with the relay mode. Data cooperation by , however, will consume energy from and it is not directly obvious which method will result in higher sum throughput for and and lower total energy consumption. In this paper, both users take advantage of RF energy harvesting whenever the other user is transmitting. There are two cases in this scenario. When the data information from one user is intended for the other user, the concept of SWIPT may be applied as the same transmitted signal can be utilized for both data and power transfer. When the transmitted information signal from one user is not intended for the other user, the RF signal can still be harvested for energy by the other user due to the broadcasting nature of wireless channels. This process can be referred to as energy salvaging. Both cases of harvesting energy from RF signals broadcast by other users are referred to as energy cooperation (EC). In the SWIPT case, when the near user receives a signal from the far user , it may split the signal into two parts according to a PS strategy, with one part for EH and the other part for relaying the message from the far user to .
Ii-a Channel Model
Referring to the system model in Fig. 1
, there are four channels in the network. Each channel can be characterized by a complex random variablewith channel power gain , which takes into account path loss and effects due to shadowing and channel fading. For simplicity, here we only consider the distance-dependent path loss so that the channel power gain is modeled as , where is the distance from the transmitter to the receiver, a path-loss exponent and an average signal power attenuation at a reference of 1 unit of distance.
We use to denote the power gain of the uplink channel from user to , and use for the power gain of the channel from to . Throughout we assume time-division multiple access (TDMA) is used for the data transmission among the nodes, hence the channel reciprocity holds, i.e., , which we shall denote by for simplicity.
All channels in our system model are quasi-static block-fading, thus the channel gains remain constant during each transmission block , but may vary from one block to another. For convenience, unit of time is assumed without loss of generality.
Ii-B Energy Harvesting Model
Fig. 2 illustrates the energy harvesting and energy flow in a user node. The charging rate for user is modeled as with for , where denotes the rate of natural energy arriving at user and is a constant specifying charging efficiency. In what follows, we assume that 1) the charging rates are equal to the associated arriving energy rates, i.e., ; 2) the charging rates remain constant during each transmission block ; 3) ideal storage devices are used to store the harvested energy without leakage so that the harvested energy is used fully for data transmission; and 4) both user nodes are able to transmit or receive signals and harvest ambient natural energy simultaneously.
Following the setup for SWIPT techniques , the RF signals transmitted by the user nodes are used for information and energy transfer. As shown in Fig. 3, the received RF signal with power , is split into two parts: with for EH and the rest for information decoding (ID).
Two extreme circumstances are for the EC-only case with all received RF signals being used for EH, and for the DC-only case with all received RF signals being used for ID. An example of a possible scenario is that in a time duration () within a transmission block of length , an inactive user harvests from the natural energy and from RF signals with average transmission power through a channel with power gain . In such a case, we have
where is the energy harvesting efficiency .
Ii-C Transmission Models
In order to be adaptive in different environmental circumstances, we propose to choose an optimal strategy from four scenarios: 1) Both DC and EC are applied; 2) Only DC is applied - the near user relays the information from to , but no energy is harvested from RF signals; 3) Only EC is applied - all received RF signals from the other user are harvested but no data relay; 4) Neither DC nor EC occurs - each user is powered by natural energy and transmits information to directly. Moreover, in each scenario we consider two cases: Case A when transmits first and Case B when transmits first. The node that transmits later will have more energy harvested for use. In the next section, by formulating and solving optimal power and time allocation for the eight combinations of energy-data cooperation and user node transmission order, the final optimal strategy will be obtained. In this section, the data transmission and energy harvesting protocol for each scenario and case are detailed. In the rest of the paper, we use S1-A to represent case A of Scenario 1, etc.
Similar to the harvest-then-transmit protocol , for all four scenarios, we assume a possible initial time interval of length with when both users harvest natural energy before transmission.
Scenario 1: Both DC and EC are applied, see Fig. 1.
Case A: transmits first.
Each transmission block consists of four non-overlapping time intervals of lengths . In the rest of this paper, we use to denote the average transmission power by sensor node with during with . The ID and EH activities at receivers in S1-A during time intervals and are summarized in Table I. As a simple explanation of the table, during , harvests the RF energy transmitted by , hence in the column the box EH is marked with a ✓. Since is inactive, the boxes in the column and are in gray color with a backslash .
During , broadcasts RF signals to and with average power , as well as harvests energy from natural resources. The complex baseband signal transmitted by is modeled as with , which is received by and as and , given by
where the noise terms are represented by and , respectively. Due to energy causality constraints , cannot use the harvested energy during until starts, and hence the energy constraint
is imposed. On the other hand, remains inactive during and harvests energy the RF signal transmitted by as well as natural energy. Since ’s message is useless for , all received RF energy is used for EH.
During , becomes active and broadcasts RF signals to and with average power . Let be the complex baseband signal transmitted by and . The corresponding received signals at and can be expressed as
where represents the noise at . And the energy constraint at is given by
Meanwhile, splits the received RF signal into two streams with PS ratio , hence for ID and for EH, where the EH part is saved for use during .
During , uses its remaining energy to relay ’s message to . The signal transmitted by is denoted as with , which is received by as , namely
The associated energy constraint during is given by
Note that although harvests energy from natural energy during , the energy is stored for use in the next transmission block, and does not contribute to the current transmission block. It is for this reason this energy is not taken into account in the formulations of energy constraints.
Finally, in one block , the amount of data transmission achievable by and can be expressed as
respectively, where . And the amounts of data transmission achievable from to (), from to (), and from to () are given by
respectively, where and . It is worth noting that we focus on the case where the source-to-relay channel is better than the source-to-destination channel , so that is also imposed. Otherwise the relay should not play a role in the transmission from source to destination .
Case B: transmits first.
In this case, each transmission block also consists of four non-overlapping time intervals, and . Unlike Case A, however, during , transmits RF signals with average power . During , forwards ’s message with average power . During , transmits its own message with average power . Table II summarizes the ID and EH activities at receivers during and .
At the end of the transmission block, the amount of data transmitted by and , denoted by and respectively, are given by
where denote the amounts of data transmission achievable from to , from to , and from to , respectively. denotes the PS ratio at and with . The energy constraints imposed on the active users are given by
Scenario 2: Only DC is applied, see Fig. 4.
In this scenario, the RF signals received at the relay node are utilized for ID without EC, i.e., and . Therefore, the data throughputs and energy constraints for Cases A and B can readily be obtained by substituting into (3), (5) and (7) - (11). The ID and EH activities during and are summarized in Table III.
Scenario 3: Only EC is applied. See Fig. 5.
Each transmission block in this scenario consists of three non-overlapping time intervals , and . The two users broadcast their own message during and , respectively. However, all RF signals received at the relay node are utilized for EH but with no data relay. The ID and EH activities at receivers are summarized in Table IV.
Under this circumstance, the achievable data transmission and associated energy constraints in Cases A and B can be expressed by
Scenario 4: Neither DC nor EC occurs.
Here, and transmit their own messages to directly during and . The data throughputs and energy constraints in S4-A and B can readily be established by substituting into (12).
Iii Convex Problem Formulation
Given natural energy arrival rates at the two user nodes and channel conditions, we seek to determine an optimal scenario in which the distribution of time intervals , power allocations over these time intervals, and data throughputs at the user nodes are jointly optimized so as to maximize the weighted sum-throughput over the present transmission block subject to the constraints imposed for the chosen scenario. We focus on weighted sum-throughput because it helps to give more attention to the sensor node with higher priority in a practical network. An alternative goal instead of weighted sum-throughput maximization is to maximize the minimum throughput of the two users . In this section we present two sets of problem formulations. Based on the analysis and models established in Section II, the joint optimization problem is formulated for each EC-DC scenario. As will be shown below, these optimization problems turn out to be convex after a simple variable substitution, hence admit reliable algorithms. The optimal scenario can therefore be identified by evaluating the solutions obtained. Furthermore, based on a second-order approximation of the two-variable “logarithmic perspective”, we derive a set of approximate problem formulations. The simplified quadratic formulations remain convex. Both problem formulations, original and approximate, have fast solutions which are developed in Section IV.
Iii-a Convex formulations based on the models from Section II
(1) Weighted sum-throughput maximization
The objective here is to maximize the weighted sum throughput over one transmission block with given throughput weights , by jointly optimizing power and time allocations for each of the four transmission scenarios. The sensor node with larger weight has higher priority.
In Scenario S1, the design variables are time allocation , power allocations , and throughputs , . Based on the model for S1 in Section II, the problems at hand for S1-A and B can be formulated, respectively, as
We note that the objective functions in (13) and (14) are in the form which is not concave with respect to and . However, by using variable substitution  and considering as the design variables instead of , the objective assumes the form which is the perspective of concave function . It then follows from Section 3.2.6 of  that is also concave with respect to variables and . The same argument applies to the constraints in problems (13) and (14) that the variable change assures that all constraints involved are convex.
Also note that the constraints in (13e) and (14e) can be used to eliminate from these problems by substituting by and imposing the non-negativity constraint . In doing so, the problems in (13) and (14) are simplified convex problems as follows, where (P1-A) and (P1-B) denote “Problem for Scenario 1, Case A” and “Problem for Scenario 1, Case B”, respectively:
Similarly, the optimization problem for S3-B can be formulated as
(2) Maximization of least throughput of the two users
As an alternative goal to the sum-throughput maximization addressed above, we may consider the problem of maximizing the least throughput of the two users. Problem of this type is known as the common-throughput maximization , and is regarded as useful as it facilitates balancing the resources between the two users. Denote the individual throughputs of users and by and , respectively, and let , the common-throughput maximization problem can be formulated as
where , and are the feasible sets for throughputs , time allocation and auxiliary variable , respectively, that can be further specified in accordance with a given scenario. Regardless of the specific scenario involved, ’s in (19b) always assume the form , hence the constraints in (19b) are convex.
Iii-B Convex quadratic formulations
(1) Quadratic approximation of the perspective function
The logarithmic perspective function
encountered in both objective function and constraints is the only non-linear component in the problems formulated in Section 3.1. As such, to a large extent it determines the computational complexity in solving these problems. An effective way to handle is to build a simple local model of the function at a given iterate (). It turns out that always has a rank-1 Hessian, hence it admits a very simple convex quadratic model surrounding () as
where , and . Fig. 6 depicts the quadratic (the surface in black) in comparison with the original (the color surface) over a fairly large region and with and . Over this region, both and vary in the range [-4.2475,-0.4615], and the closeness between the two functions in terms of the normalized Frobenius norm of over grid points of the region was found to be 0.07.
(2) Convex quadratic formulations
Based on (21), convex formulations are simplified counterparts of the problems formulated in Section 3.1 that are obtained by replacing all logarithmic perspective functions involved by quadratic approximation of the form in (21). For illustration, the local quadratic formulation of problem P1-A (see (15)) at iterate () is given by
The problem in (22) is a convex quadratically constrained quadratic programming (QCQP) problem. Based on (16) - (18), the quadratic approximation also applies to other scenarios, and QCQP formulations can be constrained for (P1-A), (P1-B), and (P2-B), and QP formulations can be obtained for (P3-A), (P3-B), (P4-A) and (P4-B). It will be shown later in Section IV, that the algorithm based on approximate QCQP is much faster than that based on the original problems.
Iv Fast algorithms for optimal energy harvesting
Energy management and data transmission are supposed to be performed in real-time. As such, optimal strategies must be constructed fast enough for them to be useful. Reliable and efficient algorithms for constrained convex problems are available, yet the computer codes realizing these algorithms are not designed to take advantages offered by particular problem structures from a particular application. In this section, we first propose an enhanced Newton barrier algorithm that is tailored to solve the original problems formulated in Section III-A. Newton methods are well known for fast convergence and high solution accuracy. They are especially suited because of the moderate problem sizes encountered here. A barrier term is incorporated to take care of the constraints and convert the problem at hand into an unconstrained convex problem. A line search technique is developed to take full advantage that each Newton direction has to offer so as to accelerate the algorithm considerably. Furthermore, as an alternative approach, we build an iterative algorithm for optimal EH based on the approximate quadratic formulations described in Section III-B.
Iv-a The Newton barrier (NB) algorithm for EH
(1) The algorithm
We consider problem P1-A as a representative formulation to illustrate the technical details of the proposed algorithm. With some notation simplified, P1-A is clearly equivalent to the following convex problem which we shall examine in the rest of Section IV-A:
where , and .
The NB algorithm solves (23) by iteratively solving the unconstrained problem