Hindawi Publishing Corporation

International Journal of Distributed Sensor Networks

Volume 2015, Article ID 586289, 10 pages
http://dx.doi.org/10.1155/2015/586289

Research Article
Enhancement of the IEEE 802.11 Power Saving Mode by
Prioritized Reservations

Xiaoying Lei and Seung Hyong Rhee

Department of Electronics Convergence Engineering, Kwangwoon University, Seoul 139-701, Republic of Korea

Correspondence should be addressed to Xiaoying Lei; [email protected]

Received 8 January 2015; Accepted 27 April 2015

Academic Editor: Chunming Qiao

Copyright © 2015 X. Lei and S. H. Rhee. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The increasing demand for real-time applications in WSN has raised the requirement of protocols considering both energy efficiency
and end-to-end delay. A PSM is proposed in the IEEE 802.11 protocol to reduce the power consumptions of wireless nodes. Wireless
nodes can stay in doze mode and periodically wake up to retrieve the frames buffered in the APs. However, the 802.11 PSM is not
such energy efficiency for WSN. First, in the process of the node’s transmitting polling frames to AP, channel contentions may cause
sensor nodes to deplete power quickly. Second, the mechanism of retrieving buffered frames can be inefficient since a polling frame
is able to pick up only one data frame. Third, a prioritized service for urgent needs is not supported. In this paper, we propose
a prioritized reservation scheme to enhance the IEEE 802.11 PSM. The concept of PSCW is suggested, during which PSM sensor
nodes can retrieve the buffered frames using the reserved SPs, where the priorities of the PSM nodes are considered in scheduling
the SPs. Through analytic models and discrete simulations, we show that our proposed mechanism outperforms the existing PSM
schemes in terms of energy efficiency and prioritized services.

1. Introduction among nodes in heavy load network may lead to collisions

which cause unnecessary power consumptions [4–8]. Sec-
The energy efficiency has been an important issue in WSN ond, if an AP sends a frame to a PSM node and more frames
(Wireless Sensor Network), since the sensor nodes operate on remained in the buffer for the node, then the AP sets More
a limited battery supply [1]. In the recent years, the increasing Data field of the frame to 1. Upon receiving the frame, the
demand for real-time applications in WSN has made the QoS sensor node keeps polling the AP until it retrieves all the
(quality of service) based communication protocols a hot remaining frames. Thus, if many frames have been buffered
topic [2]. This development in WSN raises the requirement for the PSM node, it can be very inefficient that one polling is
of designing the protocols considering both energy efficiency able to retrieve only a single data frame from the AP.
and end-to-end delay. In IEEE 802.11 MAC (medium access Additionally, 802.11 PSM has no prioritized services and
control), a PSM (Power Saving Mode) has been designed thus any urgent needs cannot be supported. Reference [9]
in order to reduce the power consumption of sensor nodes. has provided a priority scheme for 802.11 PSM such that high
According to it, sensor nodes can turn off their radios during priority nodes can transmit the PS-Poll frames earlier than
inactive periods, incoming packets for a PSM sensor node low priority ones. However, the contentions among nodes still
are buffered in the central controller node, usually is an AP exist in their scheme. Moreover, low priority nodes may waste
(Access Point), and the PSM node periodically wakes up and their energy, since they keep sensing the channel while a high
checks if it has data to receive via TIM (Traffic Indication priority node retrieves a frame.
Map) in beacon frames [3]. In this paper, we propose a prioritized reservation scheme
It has been noted that the 802.11 PSM mechanism has a to enhance the IEEE 802.11 PSM. The concept of PSCW
couple of problems [4–9]. First, in order to retrieve buffered (PSM Communication Window) is suggested, during which
traffic, the PSM sensor node polls the AP with a PS-Poll PSM nodes are allowed to retrieve the buffered frames using
frame. Since PS-Poll frames can be sent using DCF (distrib- the reserved SPs (service periods). In order to mitigate the
uted coordination function) access method, the contentions contentions among PSM nodes, AP is allowed to average
2 International Journal of Distributed Sensor Networks

the numbers of contending nodes during different BIs (bea- the nodes associate with APs that cannot overhear each other,
con intervals). Also the priority of the PSM nodes is consid- their wakeup timers can probably overlap.
ered in scheduling the service periods; that is, the channel Rozner et al. [5] present NAPman, a network-assistant
access time is allocated to PSM nodes according to their power management solution in their work. They first con-
priority. duct simulations to prove that the competing background
The rest of this paper is as follows. We outline previous traffic can significantly increase the nodes’ energy consump-
works to improve the 802.11 PSM in Section 2, and our tions and decrease the network capacity due to unneces-
proposed scheme is explained in Section 3. In Section 4, sary retransmissions. And then an energy-aware scheduling
mathematical models are built to analyze the buffering delay scheme for AP is proposed, which enables the AP to man-
and average power consumption. Simulation works and the age the active periods of nodes based on their remaining
results are followed in Section 5, and finally concluding energies. For mitigation of contentions among nodes, an AP
remarks are given in Section 6. virtualization scheme is also applied, which can lead nodes
to download traffic isolated from each other. However, this
2. Related Works approach can result in long delays for the PSM nodes with
much power left.
The IEEE 802.11 PSM offers an opportunity for sensor nodes Both of the above proposals are based on the same
to preserve their powers [10]. Based on it, the nodes having approaches allowing the AP to manage the wakeup/sleep
no traffic to transmit for a certain period can enter into doze time of nodes, using the designed scheduling mechanisms.
mode. AP buffers the incoming packets for the PSM sensor Different from their approaches, He and Yuan [6] propose a
nodes, and via broadcasting a beacon frame at the beginning TDMA-based MAC protocol for alleviating the contentions
of each BI, announces which node has frames buffered [3]. among nodes. AP divides a beacon period into a number
The PSM sensor nodes wake up periodically according to of equal time slices, each of which is allocated to a single
their LIs (listen intervals) which are equal to one or several node or a group of nodes. Therefore, instead of contending
BIs, to check the TIM (Traffic Indication Map) in the beacon for channel access, node wakes up according to its allocated
frame. If a sensor node confirms that there is data buffered time slot for data retrieving. This method effectively reduces
in the AP, it will stay active to retrieve the buffered data by energy consumption of PSM nodes by removal of channel
delivering a PS-Poll frame to the AP. On receiving the PS-Poll, contention. However, if a PSM node does not wake up in its
the AP delivers the buffered data frame to the sensor node. If time slot, the slot will be wasted. Also, as all the time slots have
still more frames are buffered for the node, the AP sets the a same length without considering packet length or traffic
More Data field of the ongoing frame to 1. Upon receiving load, the allocated time slots may be inefficiently used in case
the frame, the node keeps polling the AP until it retrieves all of short packets or light traffic.
the remaining frames. Lin et al. [7] design a DeepSleep scheme to improve
In order to receive the broadcast and multicast traffic, IEEE 802.11 PSM for machine-to-machine networks deploy-
a DTIM (delivery TIM) period, which consists of fixed ing energy harvesting devices. Because the energy levels of
numbers of BIs, is defined in 802.11 PSM. All the nodes energy harvesting devices differ from time to time and also
associating with a same AP share the same DTIM period, the harvesting rates of devices vary each other, in order
and the LI of each node can be smaller or equal to the DTIM to optimize the energy expenditure to improve the overall
period. The DTIM periods are separated by the DTIM BIs, network performance, an energy-aware sleeping algorithm
a special BI during which all PSM nodes keep active, and and a high priority algorithm are applied. The low power
a DTIM beacon frame is sent by the AP. After the DTIM devices can enter into doze mode for a certain period to
beacon, the buffered broadcast and the multicast traffic will save power, and then when they wake up, they can access
be transmitted in the DTIM BI. channel with high priority, while other devices without
Because the 802.11 PSM utilizes 802.11 DCF channel energy shortage access channel under low priority.
access method, it may suffer inefficiency especially in heavy- In [8], a balanced power saving strategy is presented,
traffic networks. The contentions among nodes may cause for determining the appropriate number of active nodes in
PSM nodes to stay active for a long period of time and waste each BI based on a tradeoff between energy consumption and
their energy. Unsuccessful transmissions (collisions or error MAC service delay. Through decreasing the number of the
channel) lead the PSM nodes to retransmit several times, active nodes, contentions for channel access are mitigated,
which can deplete their powers quickly. To alleviate this and thus PSM nodes can download their buffered traffic with
problem, various solutions have been reported in previous low delay and consuming less power. The main idea of this
literatures [4–8]. Manweiler and Choudhury [4] design a approach is similar to ours; however, we can find the power
SleepWell system for the densely deployed wireless networks. consumed for idle channel and channel access is still existing
Nodes keep a sleeping window, and the APs regulate the in [8], while it is saved by channel reservation in our protocol.
sleeping windows of their associated nodes. As different APs Moreover, there is no consideration of prioritized service in
are active/inactive during nonoverlapping time windows, by this scheme.
reducing overlap nodes can download traffic uninterrupted In order to provide prioritized services, two power saving
and enter into doze mode when the channel is occupied by approaches have been proposed for 802.11e and 802.11n
other transmissions. However, this scheme is based on the standards [3]. In 802.11e, the APSD (automatic power save
assumption that APs can communicate with each other. If delivery) is defined to take advantage of QoS mechanisms in
International Journal of Distributed Sensor Networks 3

802.11e. Nodes having no traffic for delivery can enter into BIs of sensor nodes. The modified DTIM beacon is utilized
doze mode for power saving like in the 802.11 PSM. Instead to store the rescheduled information: in legacy PSM, if the
of waking up periodically, nodes can wake up for traffic AID (association ID) of the PSM node is 3, for example, then
download at any moment by transmitting trigger frames like the third byte of PVB (partial virtual bitmap) in TIM records
Null data frames or QoS data frames to an AP. As each frame whether there is data buffered. Here we use the bytes to store
carries its priority, nodes can contend for channel access the number of a BI in which a node should wake up. For
under the supported priorities. However, the APSD suffers example, if a node with AID of 3 should wake up at the second
from the drawbacks that its uncoordinated triggering scheme BI during a DTIM period, then the third byte of the DTIM
for retrieving buffered traffic can cause high delays due to will be set to 2. Because all the sensor nodes stay active in
nodes do not transmit enough trigger frames, or else too the DTIM BI for receiving the broadcast/multicast data, by
many triggers may introduce power waste [11]. Moreover, broadcasting the DTIM beacon, AP can acknowledge nodes
the priority mechanism applied in 802.11e can introduce of their wakeup information. In such way, sensor nodes wake
unfairness problem, because the contention windows of low up not according to their LIs but to the information stored in
priority nodes are usually set to large values making the nodes DTIM, and the number of contending nodes in each BI can
have less opportunities for channel access. be averaged.
IEEE 802.11n has proposed PSMP (power saving Mul- At the start of a BI, AP initiates one PSCW for transmis-
tipoll) to enhance the APSD. A multipolling mechanism sion of PS-Poll frames and another PSCW for data frames.
is implemented in AP for scheduling the transmissions of The duration of the PS-Poll PSCW depends on the number
nodes, with a consideration of the different QoS require- of PSM sensor nodes staying awake in that BI, and the whole
ments, for example, delay or bandwidth constraints. Nodes PS-Poll PSCW is split into equal time slots with a period of
wake up for traffic download according to the scheduling exchanging one PS-Poll frame. The duration of data PSCW
information broadcast by the AP. The PSMP improves both is determined based on frame lengths and traffic load of the
energy efficiency of PSM nodes and channel utilization of active PSM nodes. With a consideration of fairness, a certain
WLANs [12]. However, the overhead of the management period of channel access time during a BI can be allocated to
frames in PSMP can be very heavy, and its implementation each group.
has a high complexity. After PSCWs are initiated, AP broadcasts the related
In [9], a simple priority scheme is given to support the information via the beacon frame. When the PS-Poll PSCW
interuser QoS. The AP utilizes the newly defined PUN packet starts, the active sensor nodes transmit PS-Poll frames in
to broadcast the number of nodes in each priority level, and turn. Here, the nodes transmit PS-Polls in the order of their
the nodes with high priority can transmit PS-Poll first while AIDs, that is, a node with a larger AID transmits earlier
the low priority nodes stay active. Nodes in the same priority than the node with a smaller one. This method is reasonable
level contend to access channel according to the DCF scheme. because the PVB of the beacon TIM element has a full
Although this scheme makes the high priority nodes retrieve illustration of the state of the nodes’ buffered traffic related to
the buffered traffic with low delay, the low priority nodes have their AIDs. On receiving a PS-Poll, AP answers with a PACK
to keep on sensing the channel while other nodes retrieve which includes the start time of an allocated SP for the node.
their traffic, which can deplete their powers quickly. PACK is a slightly modified ACK frame and the details will be
Investigating previous works as described above, in presented in the next section. If a node receives a PACK and
this paper, we propose a prioritized reservation scheme to has much time until the start time, it enters into the sleep state
improve the performance of 802.11 PSM for WSN. Our immediately; otherwise, the node stays active. In case a node
proposed protocol reduces the sensor nodes’ power con- does not wake up in its PS-Poll time slot, the SP allocated for
sumptions in sensing and contentions for channel access the node will be canceled. During data PSCW, sensor nodes
through channel reservation. Also it provides a prioritized keep active during their allocated SPs for traffic download.
service for PSM nodes. We give more details on our proposal If the reserved SP is not enough for retrieving all buffered
in the next section. traffic, AP sets the More Data field in the last packet to 1,
which makes the node wake up again in next BI.
3. Prioritized Reservation in 802.11 PSM
3.2. Modification of TIM and ACK. In order to include the
3.1. Overview. In this paper, we consider a wireless network information on PSCWs, we modify TIM as in Figure 1. Two
where an AP serves as a controller, and many sensor nodes are fields, PSCW1 and PSCW2, are added so that the start times
associated with the AP for data services. For an explanation of of PS-Poll PSCW and data PSCW are recorded, respectively.
our proposed scheme and also for simplicity, we assume that PACK that is used to announce the start time of a reserved
three levels of priority are supported in the network, and each SP is depicted in Figure 2. Since the duration field of ACK
node is assigned its priority while a node having urgent need is usually set to 0, we use this field to include the start time
is with high priority. The algorithm for assigning priority to information; note here the start time for each sensor node has
nodes is outside the scope of this work. already been scheduled by the AP at the start of the BI based
In legacy PSM, nodes wake up according to their own LIs, on nodes’ priorities, using the proposed scheduling scheme
therefore some BIs may be full of contending nodes while which will be explained later. Moreover, in the proposed
others may have only few active nodes. In order to improve scheme, only when AP receives a PS-Poll frame, it answers
this drawback, we enable the AP to reschedule the wakeup with a PACK; otherwise, the normal ACK will be transmitted.
4 International Journal of Distributed Sensor Networks

Octets: 1 1 1 1 1 1–251 1 1
Element DTIM DTIM Bitmap Partial virtual
Length PSCW1 PSCW2
ID count period control bitmap

Figure 1: TIM element in proposed scheme.

Octets: 2 2 6 4 Table 1: Nodes grouped by priority.

Frame
Start time DA CRC
control Priority Group Nodes
3 P1 A1 , A2 , A3 , . . . , A7
Figure 2: PACK in proposed scheme.
2 P2 B1 , B2 , B3 , . . . , B14
1 P3 C1 , C2 , C3 , . . . , C11

3.3. SP Allocation in Data PSCW. The SP allocated to a sensor

node is based on the traffic type and load buffered in AP. If
we define PreambleTime as the time needed for transmitting status between sleep and active frequently. For example, let
the preamble signals before data and ACK frames, define us assume that there are 5 nodes in a group that have 1, 2,
DateRate and ACKRate as the transmit rate for data and 3, 4, and 5 packets queued, respectively, at AP. If the nodes
ACK frames, respectively; we can get the expected time for transmit in a short time period first fashion, then, during the
exchanging one data frame is first node’s transmission, the second node will continue to
stay awake because of the short transmission time. Although
8𝐿 𝑖 𝑘 the third node may turn into sleep, only 3 packets later it has
𝑇𝑖 𝑘 = (2 PreambleTime + + SIFS to wake up. It should be noted that frequent state changes may
DataRate
(1) result in power waste. On the other hand, if nodes transmit
ACKSize 1 in the long time period first fashion, when a previous node
+ )×( ),
ACKRate 1−𝑟 processes its transmission, the other scheduled nodes can stay
where 𝐿 𝑖 𝑘 is the frame length for the node 𝑖 in the group of in doze for certain duration, which can preserve their powers.
priority 𝑘, ACKSize is the size of the ACK frame, and 𝑟 is the
frame loss ratio. If we define 𝑛𝑖 𝑘 as the number of packets for 3.4.1. An Example. For an example in this section, we assume
node 𝑖 in group 𝑘, the expected SP for the node is a WSN consisting of 32 sensor nodes, and DTIM BI equals
4. AP groups all nodes based on their priorities as shown
SP𝑖 𝑘 exp = 𝑛𝑖 𝑘 × 𝑇𝑖 𝑘 . (2) in Table 1 and makes them wake up into different BIs. The
active nodes and PSCW distributions in each BI are shown
To accommodate the frame retransmissions, we set a factor in Figure 3, where the transmission orders of the nodes are
𝛼 for surplus channel time, which will dynamically change scheduled by the proposed scheduling policy.
based on the evaluated channel utilization in past BIs. Thus, The details of transmission process of the sensor nodes
the SP truly allocated to node 𝑖 in group 𝑘 is are explained in Figure 4, which is a detailed figure of the first
BI in Figure 3. At the beginning of the BI, after AP initiates a
SP𝑖 𝑘 = 𝛼 × SP𝑖 𝑘 exp . (3) PS-Poll PSCW and a data PSCW, the PS-Poll PSCW is divided
into a number of slices and allocated to nodes. In the figure,
The total SPs allocated for sensor nodes belonging to a group A1 , A2 , and A3 are in group P1 , B1 and B2 are in group P2 , and
cannot exceed the total channel time allocated to that group; C1 , C5 , and C8 are in group P3 , and time slots are allocated to
that is, them. AP records the information on PS-Poll PSCW and data
SP1 𝑘 + SP2 𝑘 + ⋅ ⋅ ⋅ + SP𝑖 𝑘 ⋅ ⋅ ⋅ + SP𝑛 𝑘 ≤ 𝑐𝑘 𝑇, PSCW into the beacon frame. Upon receiving the beacon,
(4) nodes get the information on their allocated time slots. In PS-
(𝑘 = 1, 2, 3) , Poll PSCW, nodes A1 , B2 , C5 , C1 , A3 , C8 , B1 , and A2 wake up
in sequence to deliver PS-Poll frames. By receiving PACKs,
where 𝑛 is the number of PSM nodes in group 𝑘, 𝑐𝑘 is the ratio they can get the start time of the allocated SPs.
of channel time for group 𝑘, and ∑𝑘 𝑐𝑘 = 1. Because the sensor nodes B2 and C1 do not wake up
during their allocated time slots, the SPs allocated for them
3.4. SP Scheduling at AP. Note that the efficiency of schedul- are canceled. During the data PSCW, when node A1 receives
ing the allocated SPs determines that of power utilization and the first data, as More Data field is set to 1, it stays active
delay. In this section, we propose two scheduling policies. (1) continually. After the second packet is received, SP for A1
Arrange the channel time allocated to groups according to the terminates, and A1 enters into the sleep state until the next
group priority. (2) In each group, schedule all the SPs in the LI. Now nodes C5 and A3 wake up in turn to download their
order of nonincreasing length. buffered packets, and they perform the same process as A1 .
One can see that the longer the previous SP is, the longer
the next scheduled sensor node can stay in sleep. Also a short 3.5. Synchronization. Our mechanism is required for the
transmission time makes the next scheduled node change its synchronization between PSM nodes and their associated AP.
International Journal of Distributed Sensor Networks 5

BI 1 BI 2 BI 3 BI 4
Active nodes: Active nodes: Active nodes: Active nodes:
A 1 A 2 A 3 B 1 B 2 C1 C5 C8 A4 A7 B 3 B 4 B 5 C2 C3 C4 A5 A6 B 6 B 7 B 8 B 9 C6 C7 B10 B11 B12 B13 B14 C9 C10 C11

···

Beacon interval
Time
DTIM beacon interval

DTIM beacon PS-poll transmission PSCW

Beacon Data transmission PSCW

Figure 3: Nodes’ transmission order in each BI.

PVB set in TIM

PS-Poll time slots have been allocated to nodes:
AID: 8 7 6 · · · 1
A 1 B 2 C5 C1 A 3 C8 B 1 A 2

PSCW 1 PSCW 2
Beacon A1 B 2 C5 C1 A 3 A1 C5 A3

… …

PS-poll
PACK/ACK
Data

Figure 4: Transmission process.

In 802.11, the synchronization is realized by the TSF (timing sleep mode is negligible, as well as the delays and the energy
synchronization function). An AP broadcasts a beacon frame consumption incurred by state transitions. Finally, we assume
at each BI, which includes a timestamp; all the associated that packets arrive continually over time, and the service is
nodes can adjust their TSF timers according to the timestamp. assumed to be gated; that is, packets arriving in a BI are served
In this way, the nodes can keep synchronization with the AP only in or after the next BI. Note that the last assumption is
and with each other. However, this method is not perfectly reasonable since the AP has to prepare the TIM in advance
guaranteed, and loose-synchronization with AP may lead [10].
nodes to waking up at wrong time which results in packet loss
and power waste. To improve the performance of proposed
scheme, guaranteed synchronization mechanism is required 4.1. Downlink Delay. We model the average delay 𝐷 of a
for support. However, it is outside the scope of this paper. frame first, which is incurred by the power saving mechanism
at AP. Thus, the delay in this section means the sojourn time
4. Performance Analysis of a frame at the AP’s buffer. Scheduled SP for node 𝑖 in
group 𝑘 is denoted as 𝑆𝑖 𝑘 . We will explore how the position
In this section, we evaluate the performance of our proposed of 𝑆𝑖 𝑘 affects the expected delay 𝐸(𝐷). We assume that
scheme analytically. We assume that, in legacy PSM, the the incoming frames arrive at AP according to the Poisson
contention windows of AP and nodes all keep the minimum process 𝑁(𝑡) with a mean arrival rate 𝜆. Also we assume that
value (CWmin ), and after successful access to the channel, the traffic served in current BI should have arrived during last
AP or nodes transmit their data without any collisions or beacon period. Hence, each frame served in current SP will
channel errors. Further, we assume that the nodes are always experience two delays, that is, delays in last BI and current BI.
synchronized in time with AP, and the power consumption in The delay experienced in current beacon equals the sum of
6 International Journal of Distributed Sensor Networks

In legacy 802.11 PSM, the nodes and AP transmit via

26
channel contention. Based on [13], let 𝑇bo (CW𝑗 ) be the time
elapsing in backoff procedure as the contention window
22 is CW𝑗 , and let 𝑇𝑗coll be the duration in which the node
experiences a collision in its transmission. Also let 𝑝𝑓 be the
18 probability that no other nodes transmit at the same time,
and let 𝑝loss be the probability that a frame is discarded;
E(D)

14 the MAC delay conditioned to experiencing 𝑖 collisions and

successfully delivering the frame within MAX attempts is
10 𝑖+1
𝑇MAC = DIFS + ∑ 𝑇bo (CW𝑗 )
𝑗=1
6
𝑖
𝑖 (1 − 𝑝𝑓 ) × 𝑝𝑓 (8)
2 + ∑𝑇𝑗coll ,
0 20 40 60 80 100
𝑗=1 1 − 𝑝loss
S

Proposed (priority = 3) 𝑖 = 0, . . . , MAX − 1.

Proposed (priority = 2)
Proposed (priority = 1) We assume in legacy PSM, after AP received a PS-Poll from a
node successfully, AP immediately contends for the channel
Figure 5: Numerical results of 𝐸(𝐷). to transmit data without any delay. The delay of legacy PSM
can be derived as

PS-Poll PSCW duration and position of 𝑆𝑖 𝑘 , while the delay 𝑇 MAC MAC
𝑇legacy PSM = + 𝑇PS-Poll + 𝑇PS-Poll + 𝑇data , (9)
experienced in last beacon is the average delay of a Poisson 2
arrival process over the period (0, 𝑇) with the condition of
𝑁(𝑇) > 0. where 𝑇PS-Poll
MAC
+ 𝑇PS-Poll is the delay resulting from the node
Now the time spent in delivering one PS-Poll frame can transmitting PS-Poll frame and 𝑇data MA
is the delay resulting
be written as from the AP contending for channel. We find that 𝑇PS-Poll
MAC
and
PS-Poll 𝑇data are closely related to 𝑛.
MA
𝑇PS-Poll = (2 PreambleTime + SIFS +
BaseRate
(5) 4.2. Power Consumption. Using the methodology introduced
ACKSize 1 in [13], next we build an analytical model to study the power
+ )×( ),
ACKRate 1−𝑟 consumptions of PSM sensor nodes during a BI. In the IEEE
802.11 MAC, a wireless network interface can be either awake
where 𝑟 is the retransmission ratio, PS-Poll is the frame length
or in doze state. There are three different modes in the active
of the PS-Poll frame, and BaseRate is base data rate defined in
state: TX (transmit), RX (receive), and IDLE (idle), and the
802.11. The time spent in delivering a PS-Poll frame, 𝑇PS-Poll ,
power consumed in each mode is different. In the doze state,
equals to the duration of node delivering a PS-Poll frame to
two different modes have been designed, SLEEP (sleep) and
the AP, pluses the duration of SIFS (short interframe space)
OFF (power-off). The wireless network interface consumes
later, AP answering with an ACK frame. Supposing there are
much less energy in sleep state than in the awake state. In
𝑛𝑘 nodes in group 𝑘, we can get the duration of PS-Poll PSCW:
this model, we use 𝑃TX to denote the power consumed for
3 transmission, 𝑃RX for traffic reception, and 𝑃IDLE for idle state.
𝑇PSCW1 = ∑ 𝑛𝑘 × 𝑇PS-Poll . (6) Now if we let 𝑃𝐶 be the power consumed for channel
𝑘=1 contention in the legacy PSM, it can be calculated as
If we denote 𝐸(𝐷)prop as the average delay of frames suffered
CWmin
in proposed scheme, we get 𝑃𝐶 = 𝑃IDLE × (DIFS + ). (10)
2
𝑘−1 𝑖−1
𝑇 And 𝑃𝑃 which is the power consumed for exchanging a PS-
𝐸 (𝐷)prop = + 𝑇PSCW1 + ∑ 𝑆𝑚 + ∑ 𝑆𝑖 𝑘 . (7)
2 𝑚=1 𝑗=1 Poll with AP can be computed as follows:

Figure 5 shows the curves of 𝐸(𝐷) against 𝑆𝑖 𝑘 (𝑘 = 1, 2, 3), 𝑃𝑃 = 𝑃TX × 𝑇PS-Poll + 𝑃RX × 𝑇ACK + 𝑃IDLE × SIFS, (11)
where 𝑐1 = 0.3, 𝑐2 = 0.3, 𝑐3 = 0.4, 𝑇 = 0.1 seconds, and the
retransmission ratio 𝑟 is set to 0.1. The graph shows that our where 𝑇PS-Poll is the time spent in transmitting a PS-Poll
proposed scheme supports the priority mechanism, and the packet and 𝑇ACK is the time for receiving an ACK packet.
node with the high priority (priority 3) experiences the lowest Applying the methodology discussed in [9], let 𝑀 be the
delay. total number of sensor nodes in network, let 𝑇BI be the beacon
International Journal of Distributed Sensor Networks 7

interval, and let 𝜆 be the Poisson distributed packet arrival 50

rate of each node. Also let 𝛼𝑖 be the number of buffered
packets in the AP at the 𝑖th beacon interval and 𝛽𝑖 the number
of sensor nodes that have buffered packets to receive in the AP
at the 𝑖th beacon interval. Then 𝛼𝑖 and 𝛽𝑖 can be obtained as 40

Power consumption (mW)

𝛼𝑖 = 𝑀𝜆𝑇BI ,

𝛽𝑖 = 𝑀𝑃 {𝑁 (𝑇BI ) ≥ 1} = 𝑀 (1 − 𝑃 {𝑁 (𝑇BI ) = 0}) (12) 30

= 𝑀 (1 − 𝑒−𝜆𝑇BI ) .

Among the buffered packets, the number of packets belong- 20

ing to sensor nodes in group 𝑘 and the number of the nodes
belonging to 𝑘 group are derived as follows:

𝛼𝑘,𝑖 = 𝑀𝑘 𝜆𝑇BI , 10
(13) 0.6 0.7 0.8 0.9 1
−𝜆𝑇BI
𝛽𝑘,𝑖 = 𝑀𝑘 (1 − 𝑃 {𝑁 (𝑇BI ) = 0}) = 𝑀𝑘 (1 − 𝑒 ). Traffic density

Legacy PSM Proposed (priority = 3)

As a node has to stay active until all buffered traffic is Proposed (priority = 2) Proposed (priority = 1)
retrieved, the power consumed in idle state is incurred. If we
denote the power consumption as 𝑃𝑊, it can be modeled as Figure 6: Numerical results of power consumption.
follows:
𝑃𝑊 = 𝑃IDLE × 𝑇WAIT
frame, power for transmitting one PS-Poll frame during PS-
𝛼𝑖 −1 Poll PSCW, and power for receiving all buffered data frames
1
= 𝑃IDLE × (𝑇DATA𝑘 + TPS-Poll ) × ∑𝑗 during the data PSCW:
𝛼𝑖 𝑗=1 (14)
𝛼𝑘,𝑖
𝛼 −1 𝑃prop PSM = 𝑃𝐵 + 𝑃PS-Poll Tr + 𝑃
= 𝑃IDLE × 𝑖 , 𝛽𝑘,𝑖 Data Rc
2 (18)
𝛼𝑘,𝑖 󸀠
where 𝑇WAIT is the time period in which node stays in idle. 𝑃𝐷 = 𝑃𝐵 + 𝑃𝑃 + 𝑃 + 𝑃𝑊 .
𝛽𝑘,𝑖 𝐷
which is the power consumed for receiving a data frame from
AP is expressed as in (15), where 𝑇DATA 𝑘 is the time spent in Now, we compare 𝑃prop PSM and 𝑃PSM by (18) and (16), using
receiving a data packet: the system parameters given in Table 2. We vary traffic
density from 0.6 to 1.0 to have different power consumption
𝑃𝐷 = 𝑃RX × 𝑇DATA 𝑘 + 𝑃TX × 𝑇ACK + 𝑃IDLE × SIFS. (15) values, where traffic density is denoted as
Now we can calculate 𝑃PSM , the total power consumed for traffic density
the node to receive buffered frames in legacy PSM during
one BI by (16), where 𝑃𝐵 is the power consumed for receiving (number of node) × (packet length) × (traffic rate) (19)
= .
beacon frame, 𝑃PS-Poll Tr is the power for transmitting one PS- transmission rate
Poll frame, and 𝑃Data Rc is the power for data traffic:
The result is presented in Figure 6. We find that our proposed
𝛼 scheme outperforms the legacy PSM. It is because, in our
𝑃PSM = 𝑃𝐵 + 𝑃PS-Poll Tr + 𝑖 𝑃Data Rc
𝛽𝑖 scheme, the AP arranges wakeup BIs of sensor nodes which
(16) can average traffic load of BIs. Also a PSM node can reserve SP
𝛼𝑖
= 𝑃𝐵 + (𝑃𝐶 + 𝑃𝑃 ) + 𝑃 + 𝑃𝑊. for data retrieving, and thus the power consumed for channel
𝛽𝑖 𝐷 sensing and contention is saved. On the other hand, both
PSM nodes and AP access channel according to DCF in the
Next, let us model 𝑃prop PSM , the total power consumed in
legacy PSM, and the power consumed in channel contentions
proposed scheme during one BI. During data PSCW, every
rises quickly as the traffic load increasing.
node wakes up a little earlier (here we assume a period of
SIFS) before its allocated SP; thus, the power consumed in
idle mode can be calculated as 5. Simulations
𝑃𝑤󸀠 = 𝑃IDLE × SIFS. (17) In this section, simulations are performed to verify the effi-
ciency of our proposed scheme using the OPNET simulator.
Combining (11), (13), (15), and (17), 𝑃prop PSM can be expressed The performance of the proposed scheme is compared with
as (18), which is the sum of power for receiving one beacon that of the power management scheme which considers
8 International Journal of Distributed Sensor Networks

Table 2: System parameters. 60

Parameter Nodes
Packet size 1000 bytes 50
Transmit rate 11 Mbps

Power consumption (mW)

Beacon interval time 100 ms
DIFS 128 𝜇s 40
SIFS 28 𝜇s
PS-Poll 24 bytes
CWmin 32 30
𝑃RX 1.4 mw
𝑃TX 1.65 mw
20
𝑃IDLE 1.15 mw

10
interuser QoS [9] and with the legacy PSM. The simulated 0.6 0.7 0.8 0.9 1
results will be also compared with the numerical results Traffic density
obtained in the previous section.
Legacy PSM Analysis (priority = 3)
Analysis (priority = 2) Analysis (priority = 1)
5.1. Environment and Parameters. The network topology Proposed (priority = 3) Proposed (priority = 2)
consists of an AP and 50 sensor nodes, among which 20 nodes Proposed (priority = 1) Interuser (priority = 3)
have priority 3, another 20 have priority 2, and the others have Interuser (priority = 2) Interuser (priority = 1)
priority 1. The duration of BI is set to 100 ms, and in each
BI half the channel time is allocated to PSM nodes. And the Figure 7: Power consumption versus traffic density.
portions of channel time allocated to each group are set by 𝑐1 ,
𝑐2 , and 𝑐3 ; here 𝑐1 = 0.3, 𝑐2 = 0.3, and 𝑐3 = 0.4. Note that,
130
in proposed scheme, the values of 𝑐1 , 𝑐2 , and 𝑐3 can affect the
positions of allocated SPs within current BI; thus, these values
should be decided based on the traffic loads and the number
of nodes in each group. 110
The LIs of all nodes are set to 1, and the wireless protocols
are configured according to 802.11b standards. The traffics
Delay (ms)

arrived at AP are UDP (User Datagram Protocol) flows with

a fixed packet size, and frame arrivals are based on Poisson 90
process. The power model used in the simulation and other
details on the parameters are the same in Table 2.
70
5.2. Results. In the first simulation, we use various values
of the frame arrival rates, in order to get different values
of traffic density. Figure 7 presents the power consumptions
of different schemes. We find the simulation and analysis 50
0.6 0.7 0.8 0.9 1
results match each other quite well. Comparing to legacy
Traffic density
and interuser QoS PSM scheme, proposed scheme reduces
power consumption effectively. The decrement of power Interuser (priority = 1) Interuser (priority = 2)
consumption results from the reduction of idle channel and Interuser (priority = 3) Proposed (priority = 3)
the removal of contentions. In addition, when the sensor Proposed (priority = 2) Proposed (priority = 1)
node with a high priority retrieves traffic, the node with a Legacy PSM
low priority can stay in sleep mode; therefore, the power Figure 8: Delay versus traffic density.
consumptions of nodes with different priorities have a little
difference. However, in the interuser QoS scheme, a node
with a low priority has to sense the channel during other
nodes’ transmissions, which causes an energy waste. seriously, and the performances of both the legacy PSM and
Figure 8 depicts the delay of the three methods, that is, interuser QoS scheme degrade. Since the proposed method is
the legacy PSM, interuser QoS, and our proposed scheme, a scheduling based scheme, the network load does not affect
while varying the traffic density. One can see that, in a the performance much. The legacy PSM and interuser QoS
low density situation, the interuser QoS scheme performs suffer from high delays in a heavy load situation because both
much better than the other methods. However, as the traffic methods cause the backoff delay and retransmission delay.
density increases, the contentions among nodes increase We find that our proposed scheme enables nodes belonging
International Journal of Distributed Sensor Networks 9

more efficiently. Also compared to the interuser QoS scheme

30
[9], our scheme let each node retrieve the traffic at their
26
allocated time slot. This solution enables a low-priority node
to stay in sleep state while other nodes retrieve traffic, which
22 can save the power consumption of the node. Using analytic
models and discrete simulations, we can conclude that our
Delay (ms)

18 proposed scheme effectively reduces power consumptions

while supporting the nodes’ priorities and outperforms both
14 the legacy PSM and other power saving protocols.

10 Conflict of Interests
6 The authors declare that there is no conflict of interests
regarding the publication of this paper.
2
0 20 40 60 80 100
Position of scheduled SP in a BI (ms) Acknowledgments
Simulation (priority = 3) Simulation (priority = 2) This work was supported by Basic Science Research Program
Simulation (priority = 1) Analysis (priority = 3) through the National Research Foundation of Korea (NRF)
Analysis (priority = 2) Analysis (priority = 1)
funded by the Ministry of Education, Science and Technology
Figure 9: Queue buffer delay. (2013008855) and in part by the Research Grant of Kwang-
woon University in 2014.

to different groups to have different delays; that is, the delay References
of a sensor node with a high priority is lower than that of the
[1] I. Bekmezci and F. Alagöz, “Energy efficient, delay sensitive,
node with a low priority.
fault tolerant wireless sensor network for military monitoring,”
Another simulation is performed to study how the posi- International Journal of Distributed Sensor Networks, vol. 5, no.
tion of a scheduled SP affects the performance of proposed 6, pp. 729–747, 2009.
scheme, in comparison with analytical results. We use the
[2] J. Ben-Othman and B. Yahya, “Energy efficient and QoS based
simulation scenario given in [6]. Downlink streams are set routing protocol for wireless sensor networks,” Journal of
up from AP to three PSM nodes with different priorities, and Parallel and Distributed Computing, vol. 70, no. 8, pp. 849–857,
the positions of the scheduled SPs within a beacon period 2010.
are varied. We find from Figure 9 that the simulation result [3] IEEE 802.11-2012. Part 11: Wireless LAN medium access control
matches the analysis result quite well. The delay grows as the (MAC) and physical layer (PHY) specifications, IEEE, 2012.
scheduled position moves toward back, and a node with a
[4] J. Manweiler and R. R. Choudhury, “Avoiding the rush hours:
high priority has a low delay. The simulation results validate
WiFi energy management via traffic isolation,” in Proceedings
the analytic model developed in the previous section. of the 9th International Conference on Mobile Systems, Applica-
tions, and Services (MobiSys ’11), pp. 253–266, ACM, July 2011.
6. Conclusions [5] E. Rozner, V. Navda, R. Ramjee, and S. Rayanchu, “NAPman:
network-assisted power management for WiFi devices,” in
We have designed a prioritized reservation scheme for Proceedings of the 8th Annual International Conference on
enhancing the performance of the IEEE 802.11 PSM. By Mobile Systems, Applications and Services (MobiSys '10), pp. 91–
rescheduling the wakeup BIs of PSM nodes, the number of 105, San Francisco, Calif, USA, June 2010.
contending nodes during each BI can be averaged. As the [6] Y. He and R. Yuan, “A novel scheduled power saving mechanism
PSM nodes can reserve SPs for retrieving the buffered traffic for 802.11 wireless LANs,” IEEE Transactions on Mobile Comput-
by PS-Poll frames, the overhead caused by the inefficient PS- ing, vol. 8, no. 10, pp. 1368–1383, 2009.
Poll mechanism of the legacy PSM is reduced. Moreover, a [7] H.-H. Lin, H.-Y. Wei, and R. Vannithamby, “DeepSleep:
prioritized scheduling method with consideration of fairness IEEE 802.11 enhancement for energy-harvesting Machine-to-
is adopted to arrange the orders of allocated SPs, and thus Machine communications,” in Proceedings of the IEEE Global
the nodes with a high priority can retrieve their traffic Communications Conference (GLOBECOM ’12), pp. 5231–5236,
earlier than the low priority nodes, while the low priority IEEE, December 2012.
nodes allocated their own channel access time. Compared to [8] D. Jung, R. Kim, and H. Lim, “Power-saving strategy for
the scheduled scheme in [6], our proposed method allows balancing energy and delay performance in WLANs,” Computer
the node really active in current BI to reserve channel, Communications, vol. 50, pp. 3–9, 2014.
which avoids long period of empty channel. And compared [9] S.-W. Kwon and D.-H. Cho, “Efficient power management
to the power saving strategy proposed in [8], our scheme scheme considering inter-user QoS in wireless LAN,” in Pro-
eliminates the power consumed in idle channel and channel ceedings of the IEEE 64th Vehicular Technology Conference
contentions by channel reservation, which can save power (VTC-Fall ’06), pp. 1–5, Montreal, Canada, September 2006.
10 International Journal of Distributed Sensor Networks

[10] J. Lee, C. Rosenberg, and E. K. P. Chong, “Energy efficient

schedulers in wireless networks: design and optimization,”
Mobile Networks and Applications, vol. 11, no. 3, pp. 377–389,
2006.
[11] D. Camps Mur, X. Pérez-Costa, and S. Sallent Ribes, “An
adaptive solution for Wireless LAN distributed power saving
modes,” Computer Networks, vol. 53, no. 18, pp. 3011–3030, 2009.
[12] S.-L. Tsao and C.-H. Huang, “A survey of energy efficient MAC
protocols for IEEE 802.11 WLAN,” Computer Communications,
vol. 34, no. 1, pp. 54–67, 2011.
[13] G. Anastasi, M. Conti, E. Gregori, and A. Passarella, “Saving
energy in Wi-Fi hotspots through 802.11 PSM: an analytical
model,” Wireless Networks, vol. 14, no. 6, pp. 745–768, 2008.