Reliability and Delay Analysis of Slotted Anycast Multi-Hop Wireless Networks Targeting Dense Traffic IoT Applications
Summary (2 min read)
I. INTRODUCTION
- The analytical study proposed in [3] incorporates joint sleep and contention control guaranteeing throughput and SINR requirements for extending network life time.
- The authors analysis concludes that node wake up rates and active state time periods have a significant impact on the performance of the network.
II. SYSTEM MODEL
- The authors consider a generic Cyber Physical Systems (CPS) wireless network scenario with four clusters as shown in Fig. 1(a) .
- After successfully receiving a packet in Idle-Listen state relay nodes jump to Active-Tx state and wait for a beacon from their next cluster (cluster-2 is the next cluster to cluster-3).
- The relay nodes that are successful in receiving a beacon within a maximum of L a slots of the Active-Tx state follow CSMA/CA flow depicted using a 3D Markov chain shown in Fig. 2 (a) with backoff stages (m), backoff counter (k) and collision retries (n) as the three dimensions.
- Idle-Listen and Active-Tx state probabilities along with the CSMA/CA model.
- With basic understanding of busy probabilities in [4] , [5] and [6] , one can drive the mathematical model in the following section.
III. MATHEMATICAL MODEL
- The authors formulate the proposed model in two stages by deriving the transition probabilities for all states in Fig. 1 (b) in the first stage and then follows the formulation of CSMA/CA model in second stage.
- An are the probabilities to enter into the first slot of Sleep state when the received packet is discarded in CSMA/CA after exceeding maximum m and n respectively.
- P SL s |CSM Asucc is the probability of the node to enter the first slot of Sleep state after the node successfully forwards the packet.
- Reliability and delay of the proposed model are derived and analyzed in the following section.
A. Reliability model
- The reliability of a relay node can be determined by deriving the failure probabilities.
- Failure can occur due to exceeding m, n and active timeout.
- Y indicates the probability of a node transitioning to next retry after successfully sensing the channel from any of the m stages shown in Fig. 2(a) .
IV. ANALYTICAL RESULTS
- The proposed anycast clustered multi-hop analytical model's accuracy is validated by emulating a scenario similar to that shown in Fig. 1(a ) which has 4 clusters with 10 nodes each.
- The proposed emulation model has the following assumptions: Congestion due to ACK and interference from other 2.45 GHz users is negligible.
- Each relay node switches among 3 different channels for Tx, Rx and beacon modes to reduce interference between nearby nodes of different clusters.
- From the figures one can infer the importance of λ in the performance of the network.
- Increase in µ w reduces average waiting time in Active-Tx state and the chances for packet being dropped because of active timeout are less.
V. CONCLUSION
- A slotted anycast model for clustered multi-hop networks with the state-wise behaviour injected into 3D Markov chain is developed and analyzed.
- Reliability and delay performance metrics are analyzed with variation in parameters such as CSMA/CA retries, number of nodes, wake up rate and active time for different packet arrival values, and are validated using both analytical and emulation results with less than 0.5% error.
Did you find this useful? Give us your feedback
Citations
26 citations
16 citations
Cites background from "Reliability and Delay Analysis of S..."
...11 DCF inspired many studies to analyze and optimize the performance of contention-based MAC protocols for different wireless technologies using Markov chains in the literature [14], [18]–[20]....
[...]
8 citations
Cites background from "Reliability and Delay Analysis of S..."
...Multi-hop transmission is an import part of IoT WSNs, how to collect data and transmit to sink [5, 6]....
[...]
4 citations
Cites methods from "Reliability and Delay Analysis of S..."
...For the purpose of modelling, the overhead associated with beacon handshaking process between sensor node and local gateway is negligible and was considered as a realistic assumption in most popular of earlier studies such as [5], [15], [16]....
[...]
3 citations
References
461 citations
"Reliability and Delay Analysis of S..." refers background or methods in this paper
...With basic understanding of busy probabilities in [4]–[6], one can drive the mathematical model in the following section....
[...]
...After getting the relation for b0,0,0, we derive α, β and τ expressions utilizing b0,0,0 in a similar methodology shown in [6]....
[...]
136 citations
"Reliability and Delay Analysis of S..." refers background or methods in this paper
...The effect of adaptive MAC parameters on single hop and multi-hop wireless sensor networks are well studied using three-dimensional Markov models in [4] and [5] respectively....
[...]
...With basic understanding of busy probabilities in [4], [5] and [6], one can drive the mathematical model in the following section....
[...]
107 citations
"Reliability and Delay Analysis of S..." refers background in this paper
...4 MAC proposed in the literature [1], [2] & [3] do not accurately capture the state-wise behaviour of a relay node with generic routing strategy for large dense networks....
[...]
...In [1], a twodimensional Markov model for IEEE 802....
[...]
73 citations
"Reliability and Delay Analysis of S..." refers background in this paper
...01, n = [0 3], μw = [2 20], La = 100, Lp = 10, Li = 100, Ls = 100....
[...]
...4 MAC with anycast routing is proposed and in [2], a three-dimensional Markov model for IEEE 802....
[...]
...04]; (c) and (d): m0 = 3, m = 4, n = 1, La = 100, μw = [2 20], λ = [0....
[...]
18 citations
"Reliability and Delay Analysis of S..." refers background or methods in this paper
...The analytical study proposed in [3] incorporates joint sleep and contention control guaranteeing throughput and SINR requirements for extending network life time....
[...]
...4 MAC proposed in the literature [1], [2] & [3] do not accurately capture the state-wise behaviour of a relay node with generic routing strategy for large dense networks....
[...]
Related Papers (5)
Frequently Asked Questions (15)
Q2. What are the future works in "Reliability and delay analysis of slotted anycast multi-hop wireless networks targeting dense traffic iot applications" ?
The proposed integrated model and the analysis can greatly aid in driving the future research in modelling of dense traffic wireless multi-hop sensor networks. Optimization of model parameters is a focus of future research.
Q3. What is the effect of on the performance of the network?
The increase in λ results in channel congestion, leading to more packet failures due to active timeout and more delay due to backoff stages.
Q4. What is the effect of a delay model on the CSMA/CA?
Time intervals between beacons observed by a relay in Active-Tx are Poisson distributed which determine the waiting time in Active-Tx state described later in the delay model of this2 letter.
Q5. What is the simplest way to capture the data?
Sensor data collected at the leaf nodes have to be routed to a network sink via intermittent random relay nodes known as anycast routing.
Q6. What is the reason for the increase in w?
Degradation in RE is valid since increase in channel congestion values (α and β) results in more packet drops and collisions due to exceeding maximum m and n.
Q7. What is the effect of CSMA/CA retries on the delay performance metrics?
Reliability and delay performance metrics are analyzed with variation in parameters such as CSMA/CA retries, number of nodes, wake up rate and active time for different packet arrival values, and are validated using both analytical and emulation results with less than 0.5% error.
Q8. What is the effect of on the average waiting time in active state?
Increase in µw reduces average waiting time in ActiveTx state and the chances for packet being dropped because of active timeout are less.
Q9. What is the function of the relay nodes in the CPS?
In Idle-Listen state, relay nodes broadcast a beacon to the predecessor cluster (cluster-3 is the predecessor to cluster-2) and wait for a packet from it.
Q10. how many times does psls in a CSMA/CA state change?
Px = exp (−λ) , Pa = exp (−λa) (1)PS|I = PSLs |I1 = P Li x (2)PS|A = PSLs |A1 = P La a (3)λa = µw ∗N (4)PS|CSMA = PSLs |CSMAm + PSLs |CSMAn+PSLs |CSMAsucc = 1 (5)PS0 = PS0P Li x +PS0(1−P Li x )P
Q11. What is the effect of CSMA/CA on RE and Dtotal?
Fig. 2(b) and Fig. 2(c) plots RE and Dtotal versus µw respectively for 4 different values of n. RE and Dtotal are observed to be increasing and decreasing respectively with increase in µw as the average waiting time to receive a beacon in active state decreases and failures in active state due to active timeout are reduced.
Q12. What is the probability of a node to stay in the first slot of Sleep state?
Both node and cluster probabilities (α, β, τ and Pc) are in close agreement with less than 5% deviation due to equal number of nodes in each cluster and random selection of relay nodes.
Q13. what is the probability of a node to stay in the first slot of sleep state?
La a +PS|CSMAb0,0,0 (6)In Eq. (5), PSLs |CSMAm and PSLs |CSMAn are the probabilities to enter into the first slot of Sleep state when the received packet is discarded in CSMA/CA after exceeding maximum m and n respectively.
Q14. What is the effect of CSMA/CA on relay nodes?
The relay nodes that are successful in receiving a beacon within a maximum of La slots of the ActiveTx state follow CSMA/CA flow depicted using a 3D Markov chain shown in Fig. 2(a) with backoff stages (m), backoff counter (k) and collision retries (n) as the three dimensions.
Q15. What is the reliability of the proposed model?
The proposed emulation model has the following assumptions: Congestion due to ACK and interference from other 2.45 GHz users is negligible.