Pdf Fibonacci Increment Backoff Algorithm For Mac

Keywords-Backoff Algorithm, MANET, Fibonacci, Logrithmic. Defined the linear increase and decrease in the algorithm so that the. [1] IEEE Std 8, Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY).

Pdf

The IEEE 802.15.4 is a standard for Wireless Personal Area Network (PAN) that supports low data rate, low cost, low complexity and low power consumption applications. The CSMA/CA algorithm of the IEEE 802.15.4 MAC layer employs the Binary Exponential Backoff (BEB) function to compute the backoff delay for each node.

Pdf Fibonacci Increment Backoff Algorithm For Machine

Pdf Fibonacci Increment Backoff Algorithm For Mac

Using BEB function, it is possible that two or more nodes may collide if they choose the same backoff exponent value. Consequently this will increase collision and network contention level which will degrade the network overall performance.

To overcome this problem, this paper proposes a Fibonacci Backoff (FIB) function to compute the backoff interval. In FIB, each node shall wait for an incremental backoff periods as they need to access the channel.

The performance of FIB algorithm is compared against the BEB function. Beaconless modes and supports both star and peer to peer network topologies 2. In beacon enabled mode the central coordinator is responsible for sending beacons in order to synchronize nodes 2.Two successive beacons contain between them what is called a superframe 2. This superframe is divided into 16 active period time slots along with an optional inactive period in which all nodes can sleep together in the case that sleep mode is enabled.

The active period in turn consists of a Contention Access Period (CAP) and an optional Contention Free Period (CFP) 2. In order to avoid all nodes transmit at the same time, a commonly followed approach in CAP is the slotted Carrier Sense Multiple Access/ Collision Avoidance (CSMA/CA) mechanism through which nodes. 1 where BE represents the backoff exponent which is needed to find the period through which a node should stay silent aside before trying to assess the mediu m. The value of BE can be initialized to 0, 1, 2 or 3 but the default is set to 3 and in case the channel sensed busy this value shall increase to the maximum value of 5 3. This backoff delay time along with channel sensing and finally packet transmission are all happen in the backoff period which is bounded by the boundaries of each slot contained in the 16 superframe slots 3.

However, following CSMA/CA backoff strategy cannot totally avoid collisions. This is due to the fact that the backoff period is randomly chosen from the small range of 0-2. 1 and hence, there is a great possibility that more than two nodes pick up or reach the same backoff periods and as the network becomes larger in scale or nodes emit intensive and frequent traffic loads, this case is more likely to happen 3. Obviously, this situation makes those nodes which become silent for the same backoff periods detect that the availability of free channel simultaneously.

As a result, they start emitting data at the same time causing unavoidable collision. This situation leads to retransmission and consequently will degrade power consumption that is a critical constraint in WSN. This will also affect network throughput which is a result of packet losses, and hence affect net work performance 3. If we analyze the backoff mechanism carefully, we can conclude that the random nature of the binary exponential backoff (BEB) scheme is the main cause of such a problem.

Hence, unless a variation to such mechanism is done, the problem will continue to happen. However, many backoff mechanism modifications have been proposed in literature to accommodate and solve such situation, but they all still follow the BEB strategy. Previous works methods vary between changing the value of BEB parameters according to some conditions and improving the channel crier sensing (CCA) schemes 3. Obviously, the root responsible cause of the problem which is the randomness chosen period from the insufficient distributed numbers in the small range still exists. Hence, a new backoff mechanism is needed that does not depend on random chose of backoff periods.

Fibonacci

One suggested way to do so is to follow the Fibonacci Increment, a well known mathematical series that is formulated incrementally as each new subsequent number is formulated by adding the direct previous two numbers 4. The incremental behavior of Fibonacci mechanism is inspired from the Fibonacci backoff. Algorithm proposed by Manaseer 4 for IEEE 802.11 standard based networks and avoids two or more nodes to choose the same backoff period and hence avoid collision caused by BEB mechanism 4.

The following section gives an overview of some backoff schemes proposed in literature. The third section clarifies the new FIB algorithm scheme.

The fourth section illustrates some basic calculations and t heoretical scenarios. Simulation set up along with results achieved ar e illustrated in section 5 after which the paper work is concluded in section 6. Lee and Ryu (2011) 5, set the value of BE in the slotted CSMA/CA to an efficient Backoff Exponent (EBE) variable. EBE value is initialized according to the number of nodes in the PAN.

As the number of nodes joining the network decreasing, it is rational to lessen the nodes backoff delay periods. This can be accomplished by initializing EBE to the minimum value between macMinBE variable and 2 while the battery life extension (BLE) in the beacon frame is set to 1. On the contrary, collisions and channel contention shall increase as the number of nodes joining the network increases and the need for longer bac koff periods arises.

This can be accomplished by setting EBE to the average of the number of backoff delay periods by the number of nodes which are joining the network. As the value of EBE is set, it is given to BE and nodes backoff according to this value as usual. Simulation results reveal that adopting EBE increases throughput while decreases energy consumption and network load.

Khan et al (2010) 6, proposed the Improved BEB (IBEB) algorithm. IBEB minimizes the possibility that nodes may pick the same BE values and hence waiting for the same backoff time periods. In 6 nodes uses another value to calculate the backoff periods rather than choosing only BE randomly. They chose Interim Backoff (IB) value between 10% and 40% of the calculated backoff time and they also used the unit Interim Period (IP) to decrease the probability of choosing both BE and IB.

Simulation results revealed that IBEB outperforms the BEB scheme when tested on different network load and scale. (2009) 7, proposed a backoff technique named Non- Overlapping Binary Exponential B ackoff (NO-BEB) that aimed at decreasing collision level in large scale P ANs.

Posted :