Effective-Throughput Maximization for
Wireless-Powered IoT Networks with Short Packets
Jie Chen
†
, Ying-Chang Liang
†
, Xin Kang
†
, and Rui Zhang
‡
†
University of Electronic Science and Technology of China, Chengdu, P. R. China
‡
National University of Singapore, Singapore
Abstract—Internet-of-Things (IoT) is a promising technology
to connect massive machines and devices for future wireless
communications. In this paper, we study a wireless-powered IoT
network in which a hybrid access point (HAP) first transmits
power to the IoT devices wirelessly, then the devices in turn
transmit their data packets to the HAP using the harvested
energy. Due to the limited amount of energy at the devices,
the uplink data packets are in general short packets, thus
conventional resource allocation based on Shannon capacity
achieved by infinite-length packets is not optimal. In this paper,
we first define a performance metric called effective-throughput
for short packet communications (SPC), which balances the
transmission rate and the packet error rate, then establish the
effective-throughput maximization problem which optimizes the
transmission time and packet error rate for the IoT devices under
various practical constraints. Due to the non-convexity of the
formulated problem, we develop an efficient algorithm to solve it
iteratively, and provide theoretical analysis on the convergence.
Simulation results show that the proposed algorithm achieves the
similar performance as that of the exhaustive search method, and
outperforms the benchmark schemes significantly.
I. INTRODUCTION
Internet-of-Things (IoT) is a promising paradigm to support
massive connections among machines/devices in the next
generation of communications and beyond [1]–[5]. Major
applications of IoT include smart grid, healthcare, industrial
automation, and so on. There are however several challenges
to deploy IoT network in practical scenarios. One challenge
is the energy supply to IoT devices. In general, IoT devices
are powered by batteries with limited energy, and it could be
highly costly or impractical to replace the batteries regularly,
especially when the devices are deployed in harsh environment
or embedded in human bodies.
Hybrid access point (HAP) with wireless power transfer
(WPT) capability can be used to provide energy supply to
IoT devices [6], [7]. In [6], the HAP first broadcasts radio
frequency (RF) signals to charge the IoT devices (users)
and the users in turn transmit data packets to the HAP. In
[7], full duplex was used for simultaneous wireless power
transfer and information collection at the HAP. Throughput
maximization problem was studied by optimizing the transmit
time for wireless power transfer from the HAP and uplink
data transmit time from the users. These studies use Shannon
capacity as the performance metric, which implicitly assumes
infinite blocklength transmissions from the users to the HAP.
Due to the limited energies harvested at the devices, the uplink
data packets are in general short packets, thus conventional
resource allocation based on Shannon capacity achieved by
infinite-length packets is not optimal for wireless-powered IoT
networks.
In fact, it is proved in [8] that, the packet length has a
significant impact on both the maximum achievable rate and
packet error rate. In particular, the achievable rate for short
packet communications (SPC) is smaller than the Shannon
capacity, and there exists a non-zero packet error rate even
when the transmission rate is smaller than the achievable rate.
Based on the result in [8], the impact of the packet length
on the wireless-powered IoT network was further investigated
in [9]–[11]. Specifically, the effects of time allocation be-
tween the energy and information transmissions on the outage
probability and throughput were analyzed in [9]. Then, an
approximate expression was derived in [10] to characterize the
performance of error probability and latency under quasi-static
Nakagami-m fading. Besides, a large-scale Poisson network
with multiple power beacons was considered in [11], and the
energy supply probability at the transmitter and the achievable
rate of the receiver were calculated.
We notice that, most studies [9]–[11] consider the trans-
mission rate and packet error rate separately, and neglect the
joint impact of transmission rate and packet error rate. In this
paper, we take both transmission rate and packet error rate
into consideration, and adopt the effective-throughput as the
metric to balance the transmission rate and packet error rate.
Specifically, we consider a wireless-powered IoT network with
SPC, which includes an HAP and multiple users. The HAP
broadcasts RF signals to charge users in a wireless manner
and the users in turn transmit short packets to the HAP with
a practical time division multiple access (TDMA) scheme.
We first formulate a total effective-throughput maximization
problem with both the transmission time and packet error rate
of each user as variables. Due to the non-convexity of the
formulated problem, we develop an efficient algorithm based
on the principles of the block coordinate descent (BCD) algo-
rithm [12] and concave-convex procedure (CCCP) algorithm
[13], and solve the problem in an iterative manner. Besides, we
provide the convergence analysis of the proposed algorithm.
II. SYSTEM MODEL
We consider a wireless-powered IoT network shown in Fig.
1, which consists of a single-antenna HAP and K single-
antenna users, namely, U
1
, U
2
, ··· , U
K
. Each frame is
divided into two successive phases, downlink WPT phase