CN104219776A - Slotted CSMA-CA (carrier sense multiple access with collision avoidance) backoff algorithm based optimization method - Google Patents

Abstract

The invention discloses a slotted CSMA-CA (carrier sense multiple access with collision avoidance) backoff algorithm based optimization method. Priority settings are introduced to enable important command frames and short data frames of a sending node to be sent as soon as possible and data frames to be all sent at one time as soon as possible, and nodes used for channel competition and time delay caused by competition are decreased; the receiving and incorporating process of the frames is confirmed, success or not in sending is judged, and the communication process is more complete; a dynamic adjustment macMinBE method is adopted, the BE initial value is set, and adaptive adjustment is performed according to network load, idle and busy conditions of channels of previous task transmission and success or not in task sending; when the channels are busy, the value of macMinBE is increased, and probability of data collision can be reduced; when the channels are idle, the value of macMinBE is decreased, backoff time is shortened, time delay is reduced to a certain extent, power consumption is lowered, and network throughput is increased.

Description

A kind of optimization method based on time slot CSMA-CA back off algorithm

Technical field

The present invention relates to wireless communication transmission technique field, more specifically, relate to a kind of optimization method of the time slot CSMA-CA back off algorithm based on IEEE802.15.4a standard.

Background technology

CSMA-CA (csma of conflict avoidance) back off algorithm refers to when multiple radio node is attempted communicating with telegon simultaneously, judges the situation that makes fruitless efforts of channel, avoids the channel access method that data between node are collided mutually.Under different application condition and standard, its flow process also can difference to some extent.

The CSMA-CA algorithm workflow of IEEE802.15.4a standard is divided into time slot (Slotted) and non-slot (Unslotted) two kinds of modes.In actual use, a kind of mode used wherein generally can only be determined according to service condition.Here describe about the CSMA-CA algorithm under slotted mode emphatically, its flow chart as shown in Figure 1.Slotted mode sends beacon frame by telegon to carry out synchronously, and the time slot that node is vied each other subsequently, to access channel, also has the concrete service condition of specifying some time slots in beacon frame.Setting the smallest default value of BE (index avoidance) under slotted mode in agreement is macMinBE=3, and when each CSMA-CA starts, back off time needs [0,2 ^macMinBE-1] select between individual random backoff period; CW ₀the default value of (competition window length initial value) is 2, and each CSMA-CA algorithmic procedure at least needs execution twice CCA to access channel.

Although slotted mode to a certain extent can Appropriate application channel, cannot perform in time for sudden and transmission requirement that is randomness.When load is lighter, the value of macMinBE will affect the average delay of node transmission data; When heavier loads, when having multiple node all to need to send more packet simultaneously, the probability of the less generation that easily leads to a conflict of initialized setting macMinBE and re-transmission rises rapidly, and effective efficiency of transmission of channel declines, if can not be gone out by Packet Generation in time, the node needing retransmission data is caused to increase, it is fiercer to compete, vicious circle, and efficiency of transmission declines more, to cause time serious that system response is blunt even paralyses, greatly constrain real-time and the throughput of network.

Summary of the invention

In order to solve the problems of the technologies described above, the object of the invention is: a kind of optimization method based on time slot CSMA-CA back off algorithm is provided, establish the time slot CSMA-CA Markov chain Mathematical Modeling that has signal priority process and dynamic adjustments parameter function, by revising its avoidance mechanism---processed is adopted to signal priority, according to the method for channel congestion situation dynamic adjustments relevant parameter, to reduce collision and the repeating transmission of data-signal, thus reduce the energy ezpenditure of network, improve real-time and the throughput of network, obtain the performance index being better than former agreement CSMA-CA algorithm.

The technical solution adopted in the present invention is: when certain node needs to send Frame or command frame, enter time slot CSMA-CA back off algorithm, high priority signals/normal signal two kinds of ranks are adopted to distinguish the rank of data, and carry out dynamic adjustments macMinBE parameter according to channel congestion situation during last transformation task and the whether successful situation of task, flow chart after its optimization as shown in Figure 2, specifically comprises following step:

(1) initialization: for high priority signals CW ₀value be set to 1, for normal signal CW ₀value be set to 2, the value of keeping out of the way times N B is set to 0; The value of CW is set to CW ₀, the value of BE is set to macMinBE;

(2) locate back-off period beginning boundary, wait for the beginning boundary of next backoff period;

(3) [0,2 ^bE-1] get random integers in interval, and postpone should the duration of a random integers back-off period time slot;

(4) perform idle-channel evaluation CCA at back-off period boundary of time slot place to detect, judge that whether present channel is idle, if busy state, then jump to step 6; If idle condition, then enter step 5;

(5) value of CW is from subtracting 1; Judge whether current C W is 0, if be 0, then represent this CSMA-CA competitive channel success and allow node to access channel, jumping to step 7; If be not 0, then jump to step 4;

(6) value of NB is set to CW from increasing 1, CW ₀, BE is set to min (BE+1, macMaxBE); Judge that whether the value of NB is more than macMaxCSMABackoffs (maximum CSMA-CA keeps out of the way number of times restriction), if do not exceeded, then jumps to step 3; Maximumly keep out of the way number of times if exceeded, then represent this CSMA-CA competitive channel failure, report an error to upper strata, enter step 10;

(7) enter transmission state, Frame or command frame are sent;

(8) after data are sent, be confirmed whether to need wait acknowledge frame to reply, if needed, wait for the acknowledgement frame representing response in limited time, time-out does not receive the confirmation frame and then thinks that current data sends unsuccessfully, reports an error, enter step 10 to upper strata; If receive acknowledgement frame in limited time or do not need the situation of confirmation of receipt frame, enter step 9;

(9) judging the current data whether also had followed by needing to send, if also have data, jumping to step 7; If do not have data to need to send, enter step 10;

(10) dynamic adjustments macMinBE parameter, CSMA-CA performs end.

Above-mentioned high priority signals refers to that the frame of transmission is command frame or the Frame being no more than 18 byte numbers, and followed by the Frame also needing to send after successfully sending data; Except above-mentioned three kinds of situations, remaining Frame is all normal signal.

In original agreement, the value of macMinBE is that user sets, and above-mentioned macMinBE can carry out dynamic self-adapting adjustment according to the success or failure situation of channel congestion situation during last transformation task and task.Dynamic regulating method is as follows: the variables L of definition channel idle degree, for to detect channel when performing CCA in last time transmission task be idle number of times (NumOfFree) with last transmission task in perform the ratio of the total degree (NumOfCCA) of CCA, namely and introduce the reference value of the channel idle upper limit (Lmax, this programme is defined as 0.7) and channel idle lower limit (Lmin, this programme is defined as 0.3); As L>=Lmax, channel ratio is more idle, and the value of setting macMinBE is from subtracting for 1 (result is not less than 0); As L≤Lmin or last transmission mission failure, illustrate that channel is busy, the value of setting macMinBE is from adding 1 (result is no more than macMaxBE); When the value of L falls between, the value of macMinBE remains unchanged.

Wherein, the initial value set by macMinBE is 3, and its dynamic adjustments scope is [1, macMaxBE], and macMaxBE's can setting range be [3,8].

The invention has the beneficial effects as follows: use this optimization method, because introduce the setting of priority, sending node is made to have important command frame and shorter Frame to send as early as possible, Frame also can disposablely be sent as early as possible, reduces the node of channel competition use and competes the time delay brought; Include the reception of acknowledgement frame in algorithm flow, judge to send success or not, make communication process more complete; Adopt dynamic conditioning macMinBE, BE initial value is arranged, according to the situation that makes fruitless efforts of channel when offered load and last transformation task, and the success or not of task carrys out Automatic adjusument: when the value of channel compared with busy macMinBE increases, can reduce the probability of data collision; When the value of channel compared with idle macMinBE reduces, shorten back off time, decrease time delay to a certain extent, reduce power consumption, improve the throughput of network.

Accompanying drawing explanation

Fig. 1 is time slot CSMA-CA algorithm flow chart under IEEE802.15.4a standard.

Fig. 2 is the time slot CSMA-CA back off algorithm flow chart after the present invention program optimizes.

The Markov chain Mathematical Modeling figure of the node that Fig. 3 builds for this programme.

The Markov chain Mathematical Modeling figure of the channel that Fig. 4 builds for this programme.

Fig. 5 uses OPNET to the comparative graph of two schemes on delay performance.

Fig. 6 uses OPNET to the comparative graph of two schemes in throughput performance.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described further:

According to the technical solution adopted in the present invention step, flow chart as shown in Figure 2 (here no longer launch), and establishes a completely newly and accurately Markov chain Mathematical Modeling, introduces signal priority process and dynamic adjustments parameter function.Suppose the star topology network configuration be made up of 1 telegon and n node at, the node wherein sending high priority packets is N ₁node that is individual, that send general data bag is N ₂individual, and when only considering that node sends information (i.e. uplink communication) to telegon, all nodes all adopt the CSMA-CA mode access channel of band time slot, both without loss of generality, are convenient to research.One of them node of Stochastic choice, adopt three-dimensional discrete joint network model it to be carried out to the foundation of Mathematical Modeling, its one-step transition probability distributed model as shown in Figure 3.Wherein, Definition Model X (t) represents the state when moment t residing for equipment, and function expression is such as formula shown in (1).

Wherein n (t) is the NB value when moment t, and scope is [0, macMaxCSMABackoffs-1]; C (t) is the number of times performing CCA when moment t, and span is { 0,1,2}; W (t) is backoff counter, and span is [0, W _j-1], wherein W _j=2 ^{bE (t)}, BE (t) is for equipment is in the BE value of moment t.Idle, Send, Fail state represent respectively equipment be in idle condition (not having packets need to send), equipment competitive channel success and enter Packet Generation state (send duration relevant with data packet length, be set to L, using time slot as linear module) and equipment competitive channel status of fail.α, β represent respectively perform first time, second time CCA time channel busy probability.The CSMA-CA algorithm basis of former agreement is optimized, sets up the Markov probability distributed model that has the CSMA-CA of signal priority process and dynamic adjustments parameter function.

If p ₀represent that node produces the probability of new data packets, shown in (2), wherein, obeying Poisson distribution arrival rate during supposing each packet is λ _q, q is node type and q ∈ { 1,2}.

$p_0=\frac{{\mathrm{\λ}}_q}{N_q}---\left(2\right)$

Fixed $b_{i,j,k}=\underset{t\→\∞}{\lim }p\{n\left(t\right)=i,c\left(t\right)=j,w\left(t\right)=k\}$ For Markov chain Stationary Distribution, i.e. b _{i, j, k}represent that { n (t)=i, c (t)=j, w (t)=k} probability of stability, for the node of the data-signal of transmission high priority, according to the rule of Markov chain, can obtain following formula (3) to node in state.

$\left\{\begin{array}{c}{b}_{i,\mathrm{1,0}}=b_{i,\mathrm{0,0}},i\∈[0,m]\\ b_{i,\mathrm{0,0}}={\mathrm{\αb}}_{i-\mathrm{1,1,0}},i\∈[1,m]\\ b_{i,0,k}=\frac{(W_i-k)b_{i,\mathrm{0,0}}}{W_i},i\∈[0,m],k\∈[0,W_i-1]\\ b_{\mathrm{idle}}=\frac{(1-p_0)b_{\mathrm{0,0,0}}}{p_0}\\ b_{\mathrm{send}}=b_{\mathrm{success}}=(1-\mathrm{\α})\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{b}_{i,\mathrm{1,0}}\\ b_{\mathrm{fail}}={\mathrm{\αb}}_{m,\mathrm{1,0}},m=\mathrm{macMaxCSMABackoffs}\end{array}\right.---\left(3\right)$

Simultaneous formula (2) ~ (3), and all state probability sums of node are 1, can in the hope of b _0,0,0value.

$b_{\mathrm{0,0,0}}=\frac{{2p}_0(1-\mathrm{\α})}{2(1-p_0)(1-\mathrm{\α})+2p_{0}L(1-\mathrm{\α})\left({1-\mathrm{\α}}^{m+1}\right)+p_0(1-{\mathrm{\α}}^{m+1})+p_0(1-\mathrm{\α})\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{W}_i{\mathrm{\α}}^i}---\left(4\right)$

Channel idle during node execution CCA, node sends the probability γ of packet ₁:

${\mathrm{\γ}}_1=b_{\mathrm{send}}=(1-\mathrm{\α})\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{b}_{i,\mathrm{1,0}}=(1-{\mathrm{\α}}^{m+1})b_{\mathrm{0,0,0}}---\left(5\right)$

During node execution CCA, channel busy obtains probability is α:

α＝1-(1-γ ₁) ^n-1 (6)

Simultaneous formula (4) ~ (6), iterative algorithm can be calculated γ ₁with the value of α.

In like manner, for the node sending normal data signals, according to the rule of Markov chain, can in the hope of b' _0,0,0value.

$b_{\mathrm{0,0,0}}^{\′}=\frac{{2p}_0(1-\mathrm{q\α})}{2(1-p_0)(1-q)+2p_{0}L(1-\mathrm{\α})\left({1-q}^{m+1}\right)+p_0(1-q^{m+1})+p_0(1-q)\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{W}_i{\mathrm{\α}}^i}---\left(7\right)$

Wherein q=alpha+beta-α β.

Channel idle when node performs twice CCA, the probability that node sends packet is γ ₂:

${\mathrm{\γ}}_2=(1-\mathrm{\β})(1-\mathrm{\α})\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{b}_{i,\mathrm{2,0}}^{\′}---\left(8\right)$

During node execution first time CCA, channel busy obtains probability is α:

α＝1-(1-γ ₂) ^n-1 (9)

During node execution second time CCA, channel busy obtains probability is β:

$\mathrm{\β}=\frac{1-{(1-{\mathrm{\γ}}_2)}^{n-1}}{2-{\left({1-\mathrm{\γ}}_2\right)}^{n-1}}---\left(10\right)$

Simultaneous formula (7) ~ (10), iterative algorithm can be calculated γ ₂, α, β value.

Owing to adopting the transmit mechanism with priority, when supposing that node sends high priority packets, send the node then back-off delay of general data bag, can not clash between these two kinds of packets in transmission, set up the Markov model of channel, as shown in Figure 4.In Fig. 4, each state transition probability of channel model is defined as follows: α ₁without any the Packet Generation of high priority after expression execution first time CCA, proceed to the probability performing second time CCA; β ₁represent the probability that after performing first time CCA, high priority packets successfully sends; f ₁represent that after performing first time CAA, high priority packets sends failed probability; α ₂without any the probability of Packet Generation after expression execution second time CCA, β ₂represent the probability that after performing second time CCA, high priority packets successfully sends; β ₃represent that after performing second time CCA, lower-priority data is bundled into the probability of merit transmission; f ₂represent that after performing second time CCA, high priority packets sends failed probability; f ₃represent that after performing second time CCA, lower-priority data bag sends failed probability; Shown in the formula that is calculated as follows (11) of above-mentioned probability.

Definition π _k, T _kbe respectively channel model state (Idle, Busy), (Idle, Idle), the probability of stability of Success1, Success2, Fail1, Fail2 and the time of staying in this state, wherein k ∈ [1,6], has following state balance equation and residence time parameter:

π ₃＝β ₁π ₁+β ₂π ₂,π ₄＝β ₃π ₂ (12)

T ₁＝T ₂＝1,T ₃＝T ₅＝N ₁,T ₄＝T ₆＝N ₂ (13)

Flow S in network is defined as node and successfully sends the time of packet and the ratio of the total time of consumption, can be obtained the network traffics formula of two kinds of different priorities nodes as shown in the formula shown in (14) and (15) by formula (12) ~ (13).

Send the meshed network flow S of high priority packets ₁for:

$S_1=\frac{N_1[(1-{\mathrm{\α}}_2){\mathrm{\β}}_1+{\mathrm{\β}}_2{\mathrm{\α}}_1]}{{\mathrm{\α}}_1[1+({\mathrm{\β}}_2+f_2)N_1+({\mathrm{\β}}_3+f_3)N_2]+\left({1-\mathrm{\α}}_2\right)[1-\left({1-\mathrm{\α}}_1\right)N_1]}---\left(14\right)$

Send the meshed network flow S of general data bag ₂for:

$S_2=\frac{N_2{\mathrm{\α}}_1{\mathrm{\β}}_3}{{\mathrm{\α}}_1[1+({\mathrm{\β}}_2+f_2)N_1+({\mathrm{\β}}_3+f_3)N_2]+\left({1-\mathrm{\α}}_2\right)[1-\left({1-\mathrm{\α}}_1\right)N_1]}---\left(15\right)$

And average access delay performance in the present invention program, successfully send the access delay formula of a packet as shown in the formula shown in (16) and (17).

$L_1=\mathrm{Slot}\×\frac{{\mathrm{\α}}_1+(1-{\mathrm{\α}}_2)(1+N_1)}{(1-{\mathrm{\α}}_2){\mathrm{\β}}_1+{\mathrm{\β}}_2{\mathrm{\α}}_1}---\left(16\right)$

$L_2=\mathrm{Slot}\×\frac{N_2[{\mathrm{\α}}_1+(1-{\mathrm{\α}}_2)(1+N_1)]}{N_2{\mathrm{\α}}_1{\mathrm{\β}}_3}---\left(17\right)$

Wherein L ₁and L ₂represent the transmission average delay of high priority packets and general data bag respectively, the duration of Slot representation unit time slot, to send the speed that packet adopts relevant with node.

Use OPNET network law to carry out emulation in performance and com-parison and analysis to the time slot CSMA-CA algorithm after time slot CSMA-CA algorithm in IEEE802.15.4a agreement and the present invention program's optimization, simulation result as shown in Figure 5 and Figure 6.

What Fig. 5 showed is that two schemes interior joint quantity is to the Different Effects of propagation delay time.When the quantity of nodes is from 5 to 35 change, visible by analysis, have employed the present invention program, transmission delay has had obvious shortening, and data promptness gets a promotion, and especially when node is less, propagation delay time shortens about 7% than original scheme.

What Fig. 6 showed is that in two schemes, business load is to the Different Effects of throughput, and throughput here refers to the average throughput in whole network.When class interval of giving out a contract for a project changes from 0.1s/packet to 1.0s/packet, visible by analysis, have employed the present invention program, when load is less, throughput performance change is little, and when load is larger, performance has had good lifting, the original scheme of its throughput ratio has improve about 33%.

More than that principle foundation of the present invention and better performance are illustrated, but the invention is not limited to described embodiment, those of ordinary skill in the art can also make all equivalents or replacement under the prerequisite without prejudice to spirit of the present invention, and these equivalent distortion or replacement are all included in the application's claim limited range.

Claims (6)

1. based on an optimization method for time slot CSMA-CA back off algorithm, it is characterized in that, comprise following step:

(1) initialization: by the competition window length initial value CW of high priority signals ₀be set to 1, the competition window length initial value CW of normal signal ₀value be set to 2; The value of keeping out of the way times N B is set to 0; Competition window CW is set to CW ₀, the value of index avoidance BE is set to macMinBE;

(2) locate back-off period beginning boundary, wait for the beginning boundary of next backoff period;

(3) [0,2 ^bE-1] get random integers in interval, and postpone should the duration of a random integers back-off period time slot;

(4) perform idle-channel evaluation CCA at back-off period boundary of time slot place to detect, judge that whether present channel is idle, if busy state, then jump to step 6; If idle condition, then enter step 5;

(5) value of CW is from subtracting 1; Judge whether current C W is 0, if be 0, then represent this CSMA-CA competitive channel success and allow node to access channel, jumping to step 7; If be not 0, then jump to step 4;

(6) value of NB is set to CW from increasing 1, CW ₀, BE is set to min (BE+1, macMaxBE); Judge whether the value of NB exceedes maximum CSMA-CA and keep out of the way number of times restriction, if do not exceeded, then jumps to step 3; Maximumly keep out of the way number of times if exceeded, then represent this CSMA-CA competitive channel failure, report an error to upper strata, enter step 10;

(7) enter transmission state, Frame or command frame are sent;

(8), after data are sent, be confirmed whether to need wait acknowledge frame to reply, if needed, the acknowledgement frame of wait-receiving mode response in limited time, time-out does not receive the confirmation frame and then thinks that current data sends unsuccessfully, reports an error, enter step 10 to upper strata; If receive acknowledgement frame in limited time or do not need the situation of confirmation of receipt frame, enter step 9;

(9) judge current whether also have followed by the data that send of needs, if also have data, jump to step 7; If do not have data to need to send, enter step 10;

(10) dynamic adjustments parameter macMinBE, CSMA-CA performs end.

2. optimization method according to claim 1, it is characterized in that, high priority signals described in step 1 refers to that the frame of transmission is command frame or the Frame being no more than 18 byte numbers, and followed by the Frame also needing to send after current successfully transmission data; Except above-mentioned three kinds of situations, remaining Frame is normal signal.

3. optimization method according to claim 1, is characterized in that, the parameter macMinBE of dynamic adjustments described in step 10, and concrete dynamic regulating method is as follows:

When channel ratio is more idle, the value of setting macMinBE will reduce, to shorten the time delay of packet; When channel ratio comparatively busy, then the value setting macMinBE will increase, and to regulate the competition window of CSMA-CA, reduce the possibility of collision.

4. the optimization method according to claim 1 or 3, is characterized in that, the initial value set by parameter macMinBE is 3, and its dynamic adjustments scope is [1, macMaxBE], and wherein macMaxBE's can setting range be [3,8].

5. the optimization method according to claim 1 or 3, it is characterized in that, the method of dynamic adjustments macMinBE is specially: the variables L introducing channel idle degree, be the ratio performing the total degree of CCA in idle number of times and last transmission task for detecting channel when performing CCA in last time transmission task, and set the reference value of channel idle upper limit Lmax and channel idle lower limit Lmin;

As L >=Lmax, the value of setting macMinBE is from subtracting 1; As L≤Lmin or when sending unsuccessfully, the value of setting macMinBE is from increasing 1; When the value of L falls between, the value of macMinBE remains unchanged.

6. optimization method according to claim 5, is characterized in that, channel idle upper limit Lmax is 0.7, and channel idle lower limit Lmin is 0.3.