Content uploaded by Itu Snigdh
Author content
All content in this area was uploaded by Itu Snigdh on Jan 16, 2019
Content may be subject to copyright.
DSERR: Delay Sensitive Energy Efficient Reliable
Routing Algorithm
D. Gosain
1
•I. Snigdh
2
•M. Sajwan
3
Springer Science+Business Media, LLC 2017
Abstract Routing strategies need to strike a balance between responsiveness and energy
efficiency. Achieving this balance poses new challenges that do not coalesce either with
infrastructure or ad hoc wireless networks performance requirements. Multiple strategies
have emerged as a workable solution to the routing problem. Either of these solutions
however cater to but one of the many constraints posed by the networks. To address this
issue we propose a delay sensitive energy efficient reliable routing (DSERR) algorithm
which achieves application specified soft delays with energy scavenging being the fore-
most concern. We try to provide soft end-to-end (e2e) delay guarantee that is proportional
to the distance between the source and destination. Our algorithm uses the per hop greedy
selection for soft real-time guarantees as it is impossible to provide hard guarantees in a
dynamic network. DSERR presents a stateless architecture providing hop by hop reliability
of data delivery to support desired delivery reliability across the sensor network.
Keywords Wireless sensor networks Energy optimization Routing
protocols Delay Greedy based
1 Introduction
Wireless ad hoc network is characterized by autonomous nodes communicating with each
other by forming a multi hop radio network and maintaining connectivity in a decentralized
manner [1]. Today there are numerous commercial and industrial applications that require
continuous monitoring and information collection pertaining to the application they cater
&I. Snigdh
itusnigdh@bitmesra.ac.in
1
Indraprastha Institute of Information Technology, Delhi, India
2
Birla Institute of Technology, Mesra, India
3
National Institute of Technology, Delhi, India
123
Wireless Pers Commun
DOI 10.1007/s11277-017-4692-3
to. Use of wired sensors, however, may increase the cost of deployment and maintain-
ability an issue. These typical scenarios confirm the importance of employing wireless
sensor networks. Hence a wireless sensor network (WSN) is used typically in environments
where running wires or cabling is cumbersome and a fast and easy to install and maintain
network is required [1]. As a WSN is a collection of distributed sensor nodes forming an
infrastructure less network without the aid of any centralized administration; it becomes
imperative for a sensor node to work in cooperation with other nodes for transmission of
data to the sink, because of limited communication ranges. The routing protocols devel-
oped for this kind of communication need to be energy efficient and scalable. Suggested
routing algorithms for sensor networks usually aim at finding energy efficient paths to
increase the network lifetime. As a result, the ‘energy reserve’ of sensors on energy
optimized paths depletes quickly, and a ‘void or hole’ is created in the region of interest
(ROI). This ‘void or hole’ renders the sensor networks incompetent in monitoring events
occurring in some regions of their (ROI). Energy conservation thus holds the prime
importance owing to the limited battery capacity of individual nodes. Henceforth, routing
algorithms must not only consider the least cost (in terms of energy) for computing
optimized routes, but also the amount of remaining energy in each sensor, thus avoiding
uneven power consumption in the network.
There are extensive works on energy constrained routing techniques till date. The
traditional data routing classification has four main categories, namely location-based, data
centric, hierarchical and multipath [2]. Also, energy efficient routing protocols can be
classified into four main schemes: network structure, communication model, topology
based and reliable routing [3].
The delay sensitive approaches include SPEED, a stateless algorithm to provide soft end
to end delay with greedy selection of energy constrained sensors [4]; RPAR that achieves
application specific delay with dynamic adaptive power transmissions and power aware
forwarding scheme [5]. In addition, EDAL achieves a significant increase on network
lifetime without violating the packet delay constraints.
Research works focus on energy efficient multipath routing algorithms which exist in
present state of the art [6–9]. Various data centric algorithms which avoid sending data to
uninterested nodes and assume a query driven model for routing include DD [10] and SPIN
[11]. However, both of these protocols do not mention the reliability of data delivery.
Other multipath routing approaches increase the delivery reliability but also cause reduced
network lifetime and capacity due to imposed overhead [12]. Hence, choosing a multi path
routing algorithm depends on the application and involves a tradeoff between several
performance parameters.
Another category of optimized routing is the hierarchical routing strategy, in which
clusters of sensors are formed. Each cluster has a cluster head which has a responsibility to
collect data from sensors of its cluster. Popular algorithms focusing on the clustering
efficiency are LEACH [13], TEEN [14], APTEEN [15], and PEGASIS [16]. Current
literatures present cluster-based route optimization and load-balancing protocol (ROL)
which reduces the cost and number of setup messages, and thus extends the network life.
These methods improve on the robustness of algorithms like LEACH and Mires [17]by
ensuring that each node learns multiple paths to its Cluster Head [18,19].
Further, algorithms like the simple least-time energy efficient protocol discuss one level
data aggregation to increase the lifetime of network [20]. The DEBR algorithm [21] also is
a comparable algorithm in this context. Similarly, ring routing, a novel, distributed,
energy-efficient mobile sink routing protocol, suitable for time-sensitive applications,
which aims to minimize this overhead while preserving the advantages of mobile sinks
D. Gosain et al.
123
[22]. Thus, there are numerous energy-aware routing protocols designed for wireless
sensor networks where we analyze and consider a few for further analysis rather than
presenting an exhaustive review.
Despite the availability of a wide range of routing protocols for sensor networks, there
still exists a lacuna in this domain. Majority of the protocols are based on hard constraints
on resource consumption and computational complexity of the protocols running on the
nodes of these networks. Flat routing protocols such as DD and SPIN load the network
with considerably large number of packets, as they use data diffusion. This not only
reduces throughput, but also the network lifetime. Flooding is the most basic reactive
approach for WSNs and is extremely robust where we can guarantee the packet delivery
via any of the infinitesimal paths available to the sink. However, it suffers from traffic
implosion and resource blindness along with overlapping [23].
Hierarchical protocols primarily the clustering based protocols involve higher compu-
tational complexity, as they have to periodically select cluster heads and maintain the
cluster architecture. The cluster heads are also expected to have more resources and
computational capacity in many cases. Geographical information based protocol requisites
sensor nodes to be GPS equipped to enable collecting information about their location. In
addition to being costly and cumbersome, this is not a very feasible in most of the
technological solutions. Therefore, our aim is to design a simple routing protocol, which
does not require any complex mathematical evaluations, is not hardware-intensive, is
efficient, conserves energy, and also is easy to implement.
1.1 Preliminaries
Information flow between sensor nodes and the data sink is dependent on the type of
application in terms of how data is requested and used. Likewise, the overhead is measured
in terms of bandwidth utilization, power consumption, and processing requirements on the
mobile nodes. These requirements increase the complexity of the routing design problems
as well as the protocols. The routing protocols hence inadvertently need to deliver the
highest performance in scalability, reliability, responsiveness and power efficiency in
addition to data delivery. An approximate routing strategy would therefore tend to balance
these competing needs. Hence, the basic classification of routing algorithms in an ad hoc
network usually categorized as proactive, reactive and hybrid protocols need to consider
strict power and resource constraints when applied to a WSN based on a time varying
quality of a wireless channel. These inherent channel characteristics pose further challenge
attributed to the packet loss and delay. Thus, the existing protocols for routing need to
consider the intrinsic property of wireless nodes too. To address the various design
requirements there are many routing protocols suggested in literatures. Classifying routing
protocols typically for WSNs happens to be of the following types:
Class 1 Flat network architecture in which all nodes are considered as peers. This
strategy has the minimal overhead to maintain the network infrastructure and multiple
routes for fault tolerance.
Class 2 A structure based network to achieve energy efficiency, stability and scalability.
The network organizes into clusters and the head coordinates activities on behalf of its
member nodes [24].
Class 3 The data centric approach to disseminate data within the network, uses attribute
based naming. Here, merely a source node queries for the phenomenon rather than the
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
individual sensor nodes. These include broadcasting; attribute based multicasting; geo
casting and any casting [25].
Class 4 Location based routing [26] is useful in applications where the position of the
node within the geographical coverage is relevant to the query issued by the source
node.
2 DSERR Protocol
The DSERR protocol explores four main concerns of any application catered by a WSN.
Figure 1depicts the main functionalities of the proposed protocol that it seeks to achieve.
The subsections followed in the article elaborate on each of the constraints specified in this
figure. We try to devise a scheme that balances these preliminary requirements suited
commonly to every network.
It is required of a WSN to be able to deliver the required functionality with unattended
operation for the longest possible time without sacrificing the major constraints [1], with
delay and energy having prime importance. Hard end to end delays cannot be achieved,
due to unpredictable nature of WSN. Hence, DSERR aims to achieve soft application
specified delay.
The proposed approach assumes that each sensor node maintains and stores a NbrTable
that comprises of attributes like the unique ID of each neighbour node, the Euclidean
distance between itself and neighbour nodes and the residual energy of each neighbour
node. The algorithm proceeds in three main phases in its operations.
Step i Initialize NbrTable: Each sensor transmits a beacon message using the nodes’ pre-
set transmission power. This beacon packet contains its own ID and available energy.
Every node on receiving this message updates their NbrTable and stores the transmitting
node as one of its many neighbors. At the end of the initialization phase, each sensor has an
updated NbrTable and all the sensors are aware of their corresponding neighbors and their
energy attributes.
Fig. 1 DSERR star structure
D. Gosain et al.
123
Step ii Update NbrTable: The NbrTable of each node depicts the current residual power
of its consecutive neighbors. When a sensor sends a packet, all of its neighbouring nodes
receive this data along with the current energy level of the source sensor. Hence, whenever
a sensor’s energy depletes, all NbrTables, of corresponding neighbour sensors, are updated.
Step iii Routing Decision: Sensors makes a decision of the next forwarding node by
consulting its NbrTable. Initially, an intermediate nodeset is created, which stores those
node which are geographically present in between the present node and the sink. In the
next step, a delay optimal set is created which constitutes those nodes which satisfy delay
optimality condition proposed in the algorithm. In this set, the candidate node which has
least energy consumption (or maximum residual energy) is considered as next forwarding
choice.
2.1 Energy Calculations
Sensors consume energy when they sense, receive and transmit data [23,26].The amount
of energy consumed in sensing however remains unaffected by the routing algorithm
employed. We observe that an insignificant disparity exists between the overall power
consumption with or without considering the idle and receiving modes of individual
sensors [21]. The energy consumption of sensor nodes have been considered while in
transmitting mode only, thereby enabling equivalent comparisons with the existing DEBR
algorithm [21]. According to the radio model [27] energy consumption (E) for transmitting
packets is directly proportional to the square of transmission distance (d). Whilst con-
sidering normalization of the amount of sensed data, the energy consumption model is
simplified to E =d
2
[22]. Thus our analysis assumes energy consumption to be unit less.
2.2 Data Dissemination Using DSERR
The prime concern of a multi hop WSN is to select intermediate nodes to carry on the
necessary functionality of data forwarding. Determining the candidate node out of inter-
mediate nodes suitable for data forwarding between the sources to the destination is the
principal task of the routing algorithm. Figure 2presents the usual path chosen by routing
algorithms to communicate data to the sink.
Fig. 2 Geographic forwarding
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
INITIALIZATION PHASE
For each Node ‘i’
1. Create Neighbour Table (contains following attributes, ‘NbrId’ ‘NbrDist’ ‘NbrEnergy’)
2. Initialize NbrSet (based on NbrTable)
3. Initialize Packet_Dealine variable (contains approx. time in which packet must be
delivered to sink)
4. Initialize OwnEnrgy variable to default battery energy
5. Initialize Intermediate_Node_Set based in the following condition
6. For each node j
7. If dist(i,sink) – dist(j,sink) > 0
8. Add node j to Intermediate_Node_Set
9. End if
10. End For
End For
END INITIALIZATION PHASE
The proposed DSERR algorithm first computes the forward set that comprises of the
most appropriate neighboring nodes that it can forward data as shown in Fig. 3. The
Forward Set is computed as:
•Neighbor Set of Node i
NSi = {node| distance (node, node i ) ≤R}
•Forwarding Set/Intermediate Set of Node i
FSi(Destination) = {node ∈NSi| L – L_next> 0 }
Fig. 3 Computation of the forward set
D. Gosain et al.
123
2.3 Delay Sensitivity
Figure 4illustrates the delay optimal condition incorporated and computed as in the
corresponding algorithm similar to DEBR Algorithm [21]. The soft end to end delay for
individual packet is the maximum time undertaken by the packet to reach the sink. In our
case, we assume it to be 200 ns [22].
2.4 Routing Algorithm
1. Call initialization phase
2. Node A after sensing the data checks Intermediate_Node_Set
I. For each node K in FwdSetA
II. If (dist (A, Sink)/Packet_Deadline) * delay (A, K) <= dist (A, K)
III. Store node K in Delay_Optimal_Set_A
IV. End if
V. End For
3. If (Delay_Optimal_Set_A == Null)
I. Select node with least energy from NbrTable
II. Forward packet to that node
4. Else
I. Select least energy node from Delay_Optimal_Set_A
II. Transmit packet to that node
END ROUTING LOGIC
Fig. 4 Computing delay
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
Figure 5depicts the computation of neighbor table (NT) for node 3 and the corre-
sponding delay’s incurred at each forwarding path. Node 5 further selects the optimal path
which is depicted by the green arrow while the other probable suboptimal routes are
depicted as red arrows. This decision of forwarding is pre-computed by each node and
stored in its table (NT). The table entries may be updated subject to dynamics in the route
selection attributed to route congestion or node movement or node unavailability due to
energy depletions in Fig. 6.
HOLE AVOIDANCE
1. If node B has empty Delay_Optimal_Set
I. Retransmit packet to the node A (Source node)
II. If the node A has only Node B as Delay_Optimal_Node
III. remove node B from its Delay_Optimal_Set
IV. Node A retransmits packet to its sender node
V. Else
VI. removes node B from its Delay_Optimal_Set
VII. Node A transmits packet to any another node in Delay_Optimal_Set except node B
2. Process repeats iteratively till a node forwards a packet to another node in the
Delay_Optimal_Set
END HOLE AVOIDANCE
3 Results
The subsequent figures represent the snapshots of proposed algorithms with respect to the
existing algorithms implemented in Matlab. The simulation scenario assumes a terrain size
of 100 9100 m randomly populated with 50 homogeneous nodes.
Fig. 5 Computation of delay optimal forwarding set for node 5
D. Gosain et al.
123
The scenarios adopted for simulating the proposed protocol further assumes that each
node is aware of sink location through periodic beaconing to exchange location infor-
mation. However, the beaconing rate can be assumed to be very low if the sensor nodes are
static.
Figure 7a–d depict the route or path undertaken by the message in case of the analyzed
protocols namely LEO, DSERR DD and DEBR. The figures are the snap shots of the result
obtained by simulating the four different protocols. The different colors denote the dif-
ferent paths undertaken by the routing algorithm for different iterations. Figure 7a–c
denote a path with single color which indicates that the same path is followed for the
assumed number of iterations. We see that LEO protocol unnecessarily hops between a
numbers of sensor nodes just to conform to the least energy objective criterion. The
protocols depicted in Fig. 7b, d denote lesser number of hops undertaken by the packet to
reach the sink. Hence we can infer that DEBR and DSERR seem to be better than LEO and
DD as per their route selection. DEBR and DSERR protocols perform similarly, initially.
In contrast to DEBR DSERR is still better as shown in Fig. 8a, b as it selects different
sensors with increasing iterations but former is sluggish in doing so.
The different nodes chosen for different iterations enforce a uniform depletion of energy
among nodes rather than persistent depletion of nodes in the routing path. Thus DSERR
protocol claims an increase in the average lifetime at the end of ‘n’ iterations as compared
to the DEBR algorithm. Also, DEBR does not choose the routing path intelligently so as to
avoid coverage holes, as illustrated in the adjoining Fig. 9a–c. Figure 9a shows the net-
work with the source and the sink in a terrain size of 200 9200 m and 100 nodes. We
assume multiple iterations among the same source sink pair to verify the network lifetime
as well. The source and sink are chosen so as to span a computationally long path in the
network.
Figure 9b depicts the route followed under normal conditions for both DSERR and
DEBR. Figure 9c depicts the robustness of DSERR algorithm against communication
holes or node failures in the network path.
On observing node ‘19’ in Fig. 9, we see that there is no path connecting it to the sink.
Hence the packet tracks back to the intermediate sender node ‘13’ that re-computes the
Fig. 6 Recovery of message in evidence of a hole
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
route to reach the sink. In the next iteration, it chooses a different route so as to avoid the
coverage hole as shown by the route indicated in ‘yellow’ color. DEBR on the other hand
is not able to communicate to the sink and terminates abruptly.
Figures 10 and 11 depict the average energy consumed by different nodes over 100
distinct iterations for source to sink pairs. We observe and infer that DD protocol expe-
riences the maximum energy loss as it chooses the shortest path without the energy
considerations of individual nodes. LEO, conversely chooses the path with the minimum
energy consumption without considering the shortest path and other network
Fig. 7 Routing path undertaken aLEO, bDSERR, cDD, dDEBR
D. Gosain et al.
123
characteristics. DEBR appears comparatively similar to DSERR in performance till less
number of nodes.
The DEBR protocol abruptly loses its optimality with increase in the number of nodes as
shown in Figs. 10 and 11 depicts the average energy consumption for 50 nodes at the same
random positions with varying number of message packets communicated. The intention is to
observe the robustness of the proposed algorithm against varying loads. We observe that
Fig. 7 continued
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
irrespective of the number of messages delivered to the sink the average energy consumption
is the least for proposed DSERR as compared to DD, LEO and DEBR algorithms.
Figure 12 illustrates the average end to end delay and the hop count of the analyzed
protocols for 250 nodes. As the number of hops to reach the sink is lesser for DEBR and
DSERR protocols, the incurred delay is affected and is comparatively less than the DD and
LEO protocols. The end to end delay is also dependent on the number of packets retried
and delivered which aptly represents the robustness of our proposed algorithm. From
Figs. 12 and 13 we observe that the number of hop counts is the least for our proposed
strategy DSERR and so is the corresponding Delay.
Fig. 8 Path taken for 5 iterations aDSERR, bDEBR
D. Gosain et al.
123
4 Reliability
The proposed algorithm suggests a greedy approach where at each node decision for the next
hop is taken on the basis of the preeminent optimal node amongst its neighbouring nodes. An
error free and reliable communication is attributed to the Mac layer characteristics as usual.
The algorithm assumes a time out period maintained by individual nodes after transmitting a
packet to its neighbouring node for which it waits. It listens to the medium for the stipuled
Fig. 9 a Network arrangement, bcommunication path undertaken by DEBR, croute updation in event of a
communication hole by DSERR
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
time. Meanwhile, on reception of a packet in the same duration, it checks for the destination
ID. If the ID is different from its own ID, it assumes the packet has been successfully delivered
otherwise it retransmits the packet assuming that receiver received an erroneous packet. If the
sender does not receive any packet from its previously sent node (in timeout period), it again
retransmits the packet assuming it has not reached the destination. After successful
Fig. 9 continued
Fig. 10 Energy Consumed (for varying number of nodes)
D. Gosain et al.
123
Fig. 11 Energy Consumed (for varying number of message packets)
Fig. 12 Average end to end delay
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
transmission of packet node the node enters the sleep mode to conserve more energy. Thus,
the proposed approach guarantees hop-by-hop reliability.
Figure 14 shows the first node dead statistics (FND) while Fig. 15 shows the total number
of dead nodes for the protocols considered for different sensing ranges, different node den-
sities and different number of iterations. On observing Fig. 14 we see that the first failure in
DEBR algorithm is quite early as compared to LEO and proposed DSERR algorithms. From
Fig. 15, it is clear that though LEO claims to employ least energy constraints in route
determination, all nodes fail exponentially after the 250th iteration. On the other hand, DEBR
registers only a single failed node over all the iterations. This is so because, in case of DEBR,
after the first failed node, communication disrupts and hence we fail to register any change in
the statistics. We also see that the proposed DSERR algorithm registers less than 50% node
failures for 500 iterations. Due to the limitations of communication disruption in case of
DEBR, we analyze only the FND statistics with respect to DSERR as shown in Fig. 16. For
comparing the effectiveness of DSERR algorithm in terms of the lifetime we consider the
LEO algorithm and present the results depicted in Fig. 17.
5 Conclusion
Despite the disparities in the application requirements and the objectives or the respon-
sibilities of a WSN, the main task of sensor nodes is to sense and collect data from the
target domain, process the data and transmit the information back to specific sites where
the underlying application resides. Achieving this task efficiently requires the development
of an energy efficient routing strategy to set up paths between the sensor nodes and the data
sink. The path selection must be such that the overall network lifetime is maximized or
Fig. 13 Average number of hop counts
D. Gosain et al.
123
conversely, the delay is reduced. Our approach confirms that irrespective of the increase
the number of nodes or the number of packets that flow in the network to the sink, DSERR
algorithm outperforms the other algorithms. DSERR also proves to be robust enough
against node failures as it takes into consideration chances of backtracking the route in case
of coverage holes. The other comparatively similar algorithm DEBR falls short of this hole
avoidance. Thus DSERR can be applied to varying terrains or scalable networks ensuring
reliability and prolonged lifetimes. We also confirm that the strategy performs better and is
also suited for delay sensitive applications.
Fig. 14 FND statistics
Fig. 15 Number of dead nodes for considered protocols
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
References
1. Snigdh, I., & Gupta, N. (2016). Quality of service metrics in wireless sensor networks: A survey.
Journal of The Institution of Engineers (India): Series B, 97(1), 91–96.
2. Goyal, D., & Tripathy, M. R. (2012). Routing protocols in wireless sensor networks: a survey. In 2012
second international conference on advanced computing & communication technologies (ACCT) (pp.
474–480). IEEE.
Fig. 16 FND Statistics for 500 iterations
Fig. 17 Number of dead nodes after 500 iterations
D. Gosain et al.
123
3. Pantazis, N., Nikolidakis, S. A., & Vergados, D. D. (2013). Energy-efficient routing protocols in
wireless sensor networks: A survey. IEEE Communications Surveys and Tutorials, 15(2), 551–591.
4. He, T., Stankovic, J. A., Lu, C., & Abdelzaher, T. (2003, May). SPEED: A stateless protocol for real-
time communication in sensor networks. In 23rd international conference on distributed computing
systems, 2003. Proceedings (pp. 46–55). IEEE.
5. Chipara, O., He, Z., Xing, G., Chen, Q., Wang, X., Lu, C., & Abdelzaher, T. (2006, June). Real-time
power-aware routing in sensor networks. In 14th IEEE international workshop on quality of service,
2006. IWQoS 2006 (pp. 83–92). IEEE.
6. Ganesan, D., Govindan, R., Shenker, S., & Estrin, D. (2001). Highly-resilient, energy-efficient multi-
path routing in wireless sensor networks. ACM SIGMOBILE Mobile Computing and Communications
Review, 5(4), 11–25.
7. Mueller, S., & Ghosal, D. (2005, April). Analysis of a distributed algorithm to determine multiple routes
with path diversity in ad hoc networks. In Third international symposium on modeling and optimization
in mobile, ad hoc, and wireless networks, 2005. WIOPT 2005 (pp. 277–285). IEEE.
8. Ponduru, V., Ghosal, D., & Mukherjee, B. (2004, March). A distributed coverage-preserving multipath
routing protocol in wireless sensor networks. In Proceedings of the IEEE global telecommunications
conference (Globecom’04).
9. Chen, Y., & Nasser, N. (2006, August). Energy-balancing multipath routing protocol for wireless sensor
networks. In Proceedings of the 3rd international conference on quality of service in heterogeneous
wired/wireless networks (p. 21). ACM.
10. Sohraby, K., Minoli, D., & Znati, T. (2007). Wireless sensor networks: Technology, protocols, and
applications. Wiley.
11. Heinzelman, W. R., Kulik, J., & Balakrishnan, H. (1999, August). Adaptive protocols for information
dissemination in wireless sensor networks. In Proceedings of the 5th annual ACM/IEEE international
conference on Mobile computing and networking (pp. 174–185). ACM.
12. Radi, M., Dezfouli, B., Bakar, K. A., & Lee, M. (2012). Multipath routing in wireless sensor networks:
Survey and research challenges. Sensors, 12(1), 650–685.
13. Heinzelman, W. R., Chandrakasan, A., & Balakrishnan, H. (2000, January). Energy-efficient commu-
nication protocol for wireless microsensor networks. In System sciences, 2000. In Proceedings of the
33rd annual Hawaii international conference on (p. 10). IEEE.
14. Agrawal, A. M., & Agarwal, D. P. (2001). TEEN: a protocol for enhanced efficiency in wireless sensor
networks. San Francisco: IEEE.
15. Manjeshwar, A., & Agrawal, D. P. (2002, April). APTEEN: A hybrid protocol for efficient routing and
comprehensive information retrieval in wireless sensor networks. In ipdps (p. 0195b). IEEE.
16. Akkaya, K., & Younis, M. (2005). A survey on routing protocols for wireless sensor networks. Ad Hoc
Networks, 3(3), 325–349.
17. Souto, E., Guimara
˜es, G., Vasconcelos, G., Vieira, M., Rosa, N., Ferraz, C., et al. (2006). Mires: A
publish/subscribe middleware for sensor networks. Personal and Ubiquitous Computing, 10(1), 37–44.
18. Hammoudeh, M., & Newman, R. (2015). Adaptive routing in wireless sensor networks: QoS optimi-
sation for enhanced application performance. Information Fusion, 22, 3–15.
19. Kumar, S., Ranjan, P., & Radhakrishnan, R. (2015, April). EMEEDP: Enhanced multi-hop energy
efficient distributed protocol for heterogeneous wireless sensor network. In 2015 Fifth international
conference on communication systems and network technologies (CSNT) (pp. 194–200). IEEE.
20. Misra, S., & Thomasinous, P. D. (2010). A simple, least-time, and energy-efficient routing protocol with
one-level data aggregation for wireless sensor networks. Journal of Systems and Software, 83(5),
852–860.
21. Ok, C. S., Lee, S., Mitra, P., & Kumara, S. (2009). Distributed energy balanced routing for wireless
sensor networks. Computers and Industrial Engineering, 57(1), 125–135.
22. Tunca, C., Isik, S., Donmez, M. Y., & Ersoy, C. (2015). Ring routing: An energy-efficient routing
protocol for wireless sensor networks with a mobile sink. IEEE Transactions on Mobile Computing,
14(9), 1947–1960.
23. Akyildiz, I. F., Su, W., Sankarasubramaniam, Y., & Cayirci, E. (2002). Wireless sensor networks: A
survey. Computer Networks, 38(4), 393–422.
24. Chang, J. H., & Tassiulas, L. (2004). Maximum lifetime routing in wireless sensor networks. IEEE/
ACM Transactions on Networking (TON), 12(4), 609–619.
25. Wang, Q., & Yang, W. (2007, June). Energy consumption model for power management in wireless
sensor networks. In 4th annual IEEE communications society conference on sensor, mesh and ad hoc
communications and networks, 2007. SECON’07 (pp. 142–151). IEEE.
26. Stemm, M. (1997). Measuring and reducing energy consumption of network interfaces in hand-held
devices. IEICE transactions on Communications, 80(8), 1125–1131.
DSERR: Delay Sensitive Energy Efficient Reliable Routing…
123
27. Rogers, A., David, E., & Jennings, N. R. (2005). Self-organized routing for wireless microsensor
networks. IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans, 35(3),
349–359.
D. Gosain received his B.Tech. degree in Computer Science Engi-
neering from Guru Gobind Singh Indraprastha University, Delhi and
M.Tech. (Computer Science) degree from Birla Institute of Technol-
ogy, Mesra, Ranchi. Presently he is pursuing his PhD in network
anonymity privacy and anti-censorship from Indraprastha Institute of
Information Technology, Delhi.
I. Snigdh at Itu Snigdh is currently working as an Asst. Professor in
the department of Computer Science and Engg., BIT Mesra, India. She
completed her Ph.D in the area of wireless sensor networks. She has a
post graduate degree in Software Engg in 2002 from BIT Mesra. She
completed her engineering degree in electrical engineering in the year
2000. Her area of interests include software design and testing, wire-
less networks, database management systems.
M. Sajwan received his B.Tech. degree in Computer Science Engi-
neering from Amrapali Institute of Technology, Uttrakhand Technical
University, Dehradun and M.E. (software engineering) from Birla
Institute of Technology, Mesra, Ranchi. Presently, he is pursuing his
PhD in wireless sensor networks from National Institute of Technol-
ogy, Delhi.
D. Gosain et al.
123
A preview of this full-text is provided by Springer Nature.
Content available from Wireless Personal Communications
This content is subject to copyright. Terms and conditions apply.
