1 Low-Power Timely Random Access Packet-based or Connection-based
2025-04-28
0
0
1.23MB
11 页
10玖币
侵权投诉
1
Low-Power Timely Random Access: Packet-based
or Connection-based?
Tse-Tin Chan, Member, IEEE, Jian Feng, Haoyuan Pan, Member, IEEE
Abstract—This paper studies low-power random access proto-
cols for timely status update systems with information freshness
requirements, measured by age of information (AoI). In an
extensive network, a fundamental challenge is scheduling a large
number of transmitters to access the wireless channel in a way
that achieves low network-wide AoI while consuming minimal
power. Conventional packet-based random access protocols in-
volve transmitters contending for the channel by sending their
entire data packets. When the packet duration is long, the time
wasted due to packet collisions can be significant. In contrast,
connection-based random access protocols establish connections
with the receiver before transmitting data packets. From an
information freshness perspective, there should be conditions that
favor one approach over the other. We present a comparative
study of the average AoI of packet-based and connection-based
random access protocols. Specifically, we consider frame slotted
Aloha (FSA) as a representative of packet-based random access
and design a request-then-access (RTA) protocol for connection-
based random access. Our analyses indicate that the choice
between packet-based or connection-based protocols depends
mainly on the payload size of update packets and the transmit
power budget. In particular, RTA saves power and significantly
reduces AoI, especially when the payload size is large. Overall,
our investigation offers insights into the practical design of
random access protocols for low-power timely status update
systems.
Index Terms—Age of information (AoI), information freshness,
random access.
I. INTRODUCTION
The next-generation Internet of Things (IoT) network has
been envisioned as a key enabler for emerging applications
such as connected vehicles in smart cities and cyber-physical
systems in the industry. Empowered by IoT, billions of con-
nected devices with sensing, monitoring, and communication
capabilities operate collaboratively and intelligently to enable
reliable real-time communication, interactions, and decision-
making [1], [2]. In these applications, providing fresh status
updates is of utmost importance. For example, periodic safety
message exchange in vehicular networks should be delivered
among vehicles as timely as possible to ensure safety [3].
Age of information (AoI) has been a key performance met-
ric for measuring information freshness [4]–[6]. It is defined as
the time elapsed since the generation time of the latest received
This work was supported by the Guangdong Basic and Applied Basic
Research Foundation under Grant 2021A1515012601.
T.-T. Chan is with the Department of Mathematics and Information Tech-
nology, The Education University of Hong Kong, Hong Kong SAR, China
(e-mail: tsetinchan@eduhk.hk).
J. Feng and H. Pan are with the College of Computer Science and
Software Engineering, Shenzhen University, Shenzhen, 518060, China (e-
mails: fengjian2020@email.szu.edu.cn, hypan@szu.edu.cn).
information update at the destination. Specifically, suppose
that the latest information update received by the receiver is
the update packet generated by the source at time t′, then the
instantaneous AoI at time tis t−t′. Previous studies have
shown that AoI depends on the data generation pattern and
transmission delay through the network [6]. As a result, AoI
differs significantly from conventional metrics such as delay
and latency [4]–[6] and has attracted considerable research
interest, particularly for IoT applications requiring timely data.
To achieve low network-wide AoI for large-scale wireless
IoT systems, a fundamental design challenge is scheduling
massive IoT devices to access the wireless channel, especially
when these devices are typically power-constrained, such as
tiny sensors. Scheduled access protocols generally require
centralized coordination and decision-making, which can be
restrictive for many IoT scenarios with a large number of
low-cost sensors [7]. Hence, efficient random access protocols
that operate in a distributed and decentralized manner have
been extensively studied. For example, despite its simplicity,
[8] demonstrated the effectiveness of slotted Aloha (SA) in
achieving low AoI in massive access networks. In random
access, packet collisions often occur when multiple transmit-
ters send packets concurrently due to a lack of coordination,
resulting in high AoI and power wastage. While multiple
packet replicas can be sent so that advanced interference
cancellation techniques can recover the original packets and
improve AoI performance [9], transmitting multiple replicas
leads to high power consumption. This paper aims to examine
low-power random access protocols for timely status update
systems with AoI requirements.
In conventional random access schemes, a transmitter con-
tends for the channel by sending its entire data packet, known
as packet-based random access [10], [11]. Typical packet-
based random access schemes include SA and the distributed
coordination function (DCF) in WiFi networks [12]. Since
no transmitters can successfully update when more than one
transmitter sends simultaneously, the time wasted by a failed
transmission can be significant when the data packet duration
is long. In contrast to packet-based random access, connection-
based random access allows a transmitter to send a short re-
quest frame to contend for the channel first, rather than sending
the data packet directly [11]. The transmitter can only send the
data packet if it successfully receives an acknowledgment from
the receiver. The request-to-send/clear-to-send (RTS/CTS) ac-
cess mechanism in WiFi networks is a typical example of
connection-based random access: a transmitter sends an RTS
frame and waits for a CTS frame from the receiver to establish
a connection before transmitting data packets [12]. Thanks
arXiv:2210.03962v2 [cs.NI] 13 Dec 2023
2
to the established connection, the data packet transmission
is usually collision-free. Consequently, transmission failures
involve only request frames and the wasted time could be
much shorter than that of a data packet. In other words, if
the data packet is long enough, the transmission failure time
and the wasted transmit power can be reduced, i.e., the cost
of connection establishment is relatively small.
Developing AoI-aware random access schemes for timely
status update systems that consider limited transmit power
is crucial. In particular, properly using packet-based or
connection-based random access protocols is of great practical
importance. Intuitively, there should be a critical threshold
for the data packet duration beyond which it is beneficial
to establish a connection, e.g., it should be longer than the
request frame to some extent [11]. Previous works have not
investigated the AoI-aware characterization of such a thresh-
old, and this paper aims to fill this research gap. We present
a comparative study of the average AoI of packet-based and
connection-based random access protocols in the context of
an average transmit power budget.
We first consider frame slotted Aloha (FSA) as a repre-
sentative of packet-based random access protocols [10]. In
FSA, time is divided into frames, each containing a fixed
number of time slots. If a transmitter wants to send a packet
in a frame, it randomly chooses one of the time slots to
transmit. To analyze connection-based random access, we
design a request-then-access (RTA) protocol inspired by FSA.
RTA consists of two phases. In the first phase (the request
phase), transmitters contend and send update request frames
to establish a connection with the receiver. In the second
phase (the access phase), only transmitters that successfully
contended in the request phase send update packets in a
TDMA superframe, which is collision-free.
The AoI analysis of RTA is more complicated than that
of FSA. For example, the number of transmitters entering
the access phase is random and depends on the probability
of sending an update request during the request phase. More
specifically, when the probability of sending an update request
is too high or too low, the number of transmitters entering the
access phase will be small due to high collision probability
or low request update rate. Additionally, if a large number of
transmitters enter the access phase, the duration of the TDMA
superframe will also be long. In this case, the time required
between two successful updates could also be long, which also
affects the average AoI. Therefore, the probability of sending
an update request in RTA should be carefully designed to
minimize the average AoI.
We derive the closed-form average AoI and average transmit
power consumption formulas of FSA and RTA. Our simula-
tions show that whether to use packet-based or connection-
based random access protocols mainly depends on the payload
size of update packets and the transmit power budget of trans-
mitters. When the payload size is large, RTA often outperforms
FSA because RTA dedicates a connection establishment phase
to help avoid collisions of data packets, thus saving power and
reducing AoI. When the duration of an update packet and a
request frame are comparable (e.g., 16 bytes), RTA should still
be used when the power budget of transmitters is high enough;
otherwise, FSA is a simple and effective solution.
To sum up, this paper has the following three major contri-
butions:
(1) We are the first to compare packet-based and connection-
based random access protocols in low-power IoT status
update systems with AoI requirements. Specifically, for a
given transmit power budget, we study different random
access protocols to achieve high information freshness.
(2) We use frame slotted Aloha (FSA) as the representatives
of packet-based protocols and design request-then-access
(RTA) as the connection-based protocol for theoretical
analysis. Closed-form average AoI and average transmit
power consumption formulas of different protocols are
derived. Our study serves as a guideline for comparing
packet-based and connection-based random access.
(3) We conduct comprehensive simulations to evaluate the
performances of different protocols. The simulation re-
sults reveal that the favorability of connection-based
random access depends mainly on the payload size of the
update data packets (compared with the request frames),
as well as the transmit power budget. Overall, our investi-
gations provide insights into the design of random access
protocols for low-power timely status update systems.
II. RELATED WORK
A. Age of Information (AoI)
AoI was first proposed in vehicular networks to characterize
the timeliness of safety packets [13]. After that, it has been
extensively studied under various communication and network
systems; see the monograph [5] and the references therein for
important research results. Most early studies of AoI focused
on the upper layers of the communication protocol stack, i.e.,
above the physical (PHY) and medium access control (MAC)
layers. For example, a rich literature analyzed the AoI perfor-
mance under different abstract queueing models, e.g., single-
source single-server queues [14], multiple-source single-server
queues [15], etc. To lower the network-wide AoI, age-optimal
scheduling policies among multiple transmitters are examined
in [16]–[18], with the goal of minimizing different AoI metrics
at a common receiver, such as average AoI and peak AoI [1].
Moving down to the PHY and MAC layers, considering
imperfect updating channels with interference and noise, the
average AoI was analyzed in different network topologies,
such as multi-hop [19] and multi-source [20] networks. More-
over, different age-oriented error-correction techniques were
investigated to combat the wireless impairments, e.g., auto-
matic repeat request (ARQ) [21] and channel coding [22],
[23]. These works reveal that age-optimal designs are usually
different from conventional delay/latency-optimal ones.
Energy is another crucial issue when designing AoI-aware
status update systems, especially for low-power IoT sensor
networks. For example, the AoI-energy characteristics of status
update systems when using different ARQ protocols were
discussed in [21], [24], [25]. These works showed an inherent
tradeoff between the AoI and the average energy consumption
at the transmitters, especially under the generate-at-will model.
In other words, the transmitter needs to decide whether to send
3
(or resend) a packet under the transmit power constraint [21].
Unlike previous works, we do not consider ARQ in this paper.
Instead, we study different MAC protocols in random access
channels, with the goal of reducing the average AoI given an
average transmit power budget.
B. Random Access Protocols related to AoI
In the literature, different wireless access schemes on AoI
were investigated, including scheduled access [7], [26] and
random access [9], [27], [28]. Compared with scheduled access
protocols, efficient random access protocols that operate in
a distributed and decentralized way have shown promising
results in IoT networks [8], [29]–[31]. Along this line, our
work also focuses on AoI-aware random access protocols due
to their importance and ease of implementation in distributed
networks.
There has been rich literature on the modeling and per-
formance analysis of AoI-aware random access protocols. In
particular, slotted Aloha (SA)-based protocols have received
the most attention. Despite being a relatively simple protocol,
[8] showed the AoI effectiveness of SA in massive access
networks. Moreover, successive interference cancellation (SIC)
was applied to improve the AoI performance of SA when the
number of transmitters is large (i.e., the collision probability is
high) [9], [32]. Moving beyond SA, the AoI of carrier sensing
multiple access (CSMA) and general random access protocols
were analyzed in [33] and [34], respectively. Compared with
the literature, this paper further investigates the average AoI of
frame slotted Aloha (FSA), a generalization of SA. The AoI
analysis of FSA was studied recently in [35], [36]. Specifically,
[35] investigated the effects of retransmission on the average
AoI of FSA, and [36] derived analytical expressions for the
average and variance of AoI over a typical transmission link
in Poisson bipolar and cellular networks. However, the above
works focused extensively on the packet-based random access
[11].
To the best of our knowledge, the comparison between
packet-based and connection-based random access protocols
has not been investigated in the literature. This paper designs
a request-then-access (RTA) protocol as the representative
of connection-based random access. In the IEEE 802.11
standards, the payload size of data packets should exceed
a threshold to activate the RTS/CTS mechanism so that the
connection establishment facilitates higher network throughput
[12]. Unlike conventional 802.11 networks, our work con-
siders AoI as the performance metric rather than network
throughput. We notice that a recent worok [37] studied the
AoI performance of a reservation-based random access scheme
that is similar to RTA. However, a fixed-duration access phase
was considered in [37], which possibly leads to time wastage
when most of the transmitters collide in the reservation phase
(similar to the request phase in RTA). In contrast, RTA
considers an access phase with variable duration depending
on the number of transmitters that successfully contended in
the request phase. We examine how the request phase in RTA
reduces transmission failure time and transmit power wastage
to achieve timely and low-powered state updates.
3…
1
2
4
5
6
N
Sensor Access Point (AP)
Fig. 1. A status update system with Nsensors sending update packets to the
common access point (AP) in a random access manner.
III. AGE OF INFORMATION PRELIMINARIES
A. Age of Information (AoI)
We study a timely status update system in which Nsensors
send status update packets to a common access point (AP)1,
as shown in Fig. 1. At time instant t, the instantaneous AoI
of sensor u, denoted by ∆u(t), is defined by
∆u(t) = t−Gu(t),(1)
where Gu(t)is the generation time of the latest update packet
received by the AP from sensor u[4]–[6]. A lower ∆u(t)
means a higher degree of information freshness.
With the instantaneous AoI ∆u(t), we can compute the
average AoI of sensor u. The average AoI ¯
∆uis defined as
the time average of the instantaneous AoI [4]–[6]
¯
∆u= lim
T→∞
1
TZT
0
∆u(t)dt. (2)
A low average AoI ¯
∆uindicates that the update packets from
sensor uare generally fresh over a long period. Considering
the whole system, the average AoI for all the sensors is ¯
∆ =
1
NPN
u=1 ¯
∆u.
This paper considers a collision model, where a
packet/frame is successfully received only when no other
sensors’ packets/frames are sent at the same time; otherwise,
simultaneous transmissions from multiple sensors lead to a
collision [10]. Due to packet collision, sensor umay send
more than one update packet until the next successful update.
Fig. 2 plots an example of the instantaneous AoI ∆u(t)in
which the (j−1)-th and the j-th successful updates occur
at tj−1and tj, respectively. We see that sensor usends four
update packets after the last successful update at tj−1, and
only the last one is successfully received by the AP at tj. If
an update packet is not received successfully, the instantaneous
AoI ∆u(t)continues to increase linearly.
Let Tpk denote the transmission time of an update packet.
This paper considers the generate-at-will model, where a sen-
sor can take measurements and generate a new update packet
when it has the opportunity to transmit. The instantaneous AoI
1This paper assumes that the number of sensors, N, is known. In a practical
random access network, it is difficult to obtain the exact Nat the receiver,
which requires online estimation. The estimation of Nis out of the scope of
the current work, and we refer interested readers to [38] for more details. The
impact of an inaccurate Non the AoI performance of different protocols will
be presented in Section V.
摘要：
展开>>
收起<<
1Low-PowerTimelyRandomAccess:Packet-basedorConnection-based?Tse-TinChan,Member,IEEE,JianFeng,HaoyuanPan,Member,IEEEAbstract—Thispaperstudieslow-powerrandomaccessproto-colsfortimelystatusupdatesystemswithinformationfreshnessrequirements,measuredbyageofinformation(AoI).Inanextensivenetwork,afundamenta...
声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！
相关推荐
-
【词汇变形总汇】2025高考词汇变形总汇 - 教师版VIP免费
2024-12-06 3 -
【超简37页】新课标高考英语考纲3500词汇VIP免费
2024-12-06 4 -
《高考英语3500词详解》（WORD版）VIP免费
2024-12-06 13 -
《高考英语3500词详解》VIP免费
2024-12-06 11 -
高中英语-[教师版]80天通关高考3500词汇VIP免费
2024-12-06 12 -
高中人教选修7课文逐句翻译VIP免费
2024-12-06 7 -
高中人教选修7课文原文及翻译VIP免费
2024-12-06 13 -
高中人教必修4课文逐句翻译VIP免费
2024-12-06 7 -
高中人教必修4课文原文及翻译VIP免费
2024-12-06 13 -
高考英语核心高频688词汇VIP免费
2024-12-06 10
分类：图书资源
价格：10玖币
属性：11 页
大小：1.23MB
格式：PDF
时间：2025-04-28
作者详情
相关内容
-
3-专题三 牛顿运动定律 2-教师专用试题
分类：中学教育
时间：2025-04-07
标签：无
格式：DOCX
价格：5.9 玖币
-
2-专题二 相互作用 2-教师专用试题
分类：中学教育
时间：2025-04-07
标签：无
格式：DOCX
价格：5.9 玖币
-
6-专题六 机械能 2-教师专用试题
分类：中学教育
时间：2025-04-07
标签：无
格式：DOCX
价格：5.9 玖币
-
4-专题四 曲线运动 2-教师专用试题
分类：中学教育
时间：2025-04-08
标签：无
格式：DOCX
价格：5.9 玖币
-
5-专题五 万有引力与航天 2-教师专用试题
分类：中学教育
时间：2025-04-08
标签：无
格式：DOCX
价格：5.9 玖币
渝公网安备50010702506394
