CN102413540B - Self-organizing network unicast method with combination of cognitive network coding and routing based on cognition - Google Patents

Abstract

The invention discloses a self-organizing network unicast method with combination of cognitive network coding and routing based on cognition. The method comprises the following steps: each node in a self-organizing network periodically sends a parameter state frame to carry out topology discovery and parameter perception, and determines butterfly topology and a key node; the key node respectively calculates an end-to-end time delay estimated value of a route unicast strategy and a network coding unicast strategy; the key nodes compares the end-to-end time delay estimated values of the above two unicast strategies, selects an unicast strategy with a smaller estimated value, and notifies the selected unicast strategy to a top node and a bottom node in the butterfly topology; the key node, the top node and the bottom node employ the selected unicast strategy to complete grouping transmission together. According to the invention, end-to-end time delay is reduced, respective advantages of the network coding and routing are effectively utilized, bearing capability of the self-organizing network to a unicast service is raised, and the method can be used in a unicast scene data transmission process with the butterfly topology in a wireless self-organizing network.

Description

The self-organizing network unicast method that network code based on cognitive combines with route

Technical field

The present invention relates to wireless communication field, the unicast method that particularly a kind of network code based on cognitive combines with route, is applied in clean culture contextual data transmitting procedure that wireless self-organization network has butterfly topology.

Background technology

Network code is a kind of novel data transmission scheme, by people such as R.Ahlswede, in 2000, is proposed first, and its core concept is to allow intermediate node to encode to the information from different links, as XOR or linear combination, and then sends.It has been overthrown at intermediate node place the data of receiving is processed and can not be brought this traditional viewpoint of any income, the change that has therefore brought essence to network service.The topmost advantage of network code is to reduce data transmission times by increasing single transmission amount of information, and then improves the throughput of network; Due to the theory of network coding cooperation, need coding nodes and its surrounding's node by coding-decoding operation, jointly to complete the transmission of data simultaneously, can share offered load like this, what while effectively having avoided load overweight, at coding nodes place, produced is congested.

In wireless self-organization network, to compare with traditional routing plan, the performance of network code is relevant with following factor:

1. the business in network.Because business in self-organizing network is random arrival, so the time interval that business arrives is also random.And node to carry out condition that network code obtains gain be a plurality of from different links, at least two business just can be carried out after all arriving node, and then transfer to again next-hop node after merging into a business, when a plurality of business arrive time interval of nodes when longer, node is longer in the time of carrying out waiting for before network code, has caused the decline of network throughput.When adopting traditional routing plan transmission grouping, under the heavier situation of offered load, after needing first to wait in line after a plurality of groupings arrival nodes, by node, it is carried out to timesharing transmission again.

2. transmission of radio links grouping makes mistakes.No matter adopt which kind of scheme, the object of transmitted in packets is all to allow all groupings can successfully arrive destination node separately.When Adoption Network encoding scheme transmission of data packets, a link transmission is made mistakes and can be caused a plurality of destination nodes successfully to decode, and visible network encoding scheme is higher to the reliability requirement of link.When adopting routing plan transmission grouping, a link transmission is made mistakes and only can be made a destination node cannot obtain required grouping.

3. network topology.In self-organizing network, the frequent mobile topology that causes of node constantly changes, so network code and routing plan need obtain network topological information, obtains network topological information and can increase network overhead when network size is larger.In sum, these two kinds of transmission plans of network code and route have his own strong points because application scenarios is different.

The advantage raising self-organizing network that how to make full use of network code is one of key issue of current network coding study to the bearing capacity of unicast service.Although existing coded data unicast scheme Network Based can make full use of the advantage of network code, can not be advantageously applied in wireless self-organization network, reason is as follows:

The first, existing coded data unicast scheme Network Based can not adapt to the change at random at the business interval time of advent in self-organizing network.After all arriving, a plurality of business that node will participate in encoding by the time just carry out network code, when time interval that a plurality of business arrive is larger, the time that node is waited for is longer, end-to-end time delay is larger, this has just caused the decline of network throughput, thereby the also decline rapidly of the performance of network code scheme, under this situation, the performance of route unicast scheme is better than network code scheme on the contrary.

The second, existing scheme is not considered the impact that transmission of radio links is made mistakes on destination node decoding.During Adoption Network coding unicast scheme, a link transmission is made mistakes and can be had influence on the successful decoding of a plurality of destination nodes, so when link error probability is larger and be not suitable for Adoption Network encoding scheme and transmit data.

The 3rd, existing scheme is not considered the dynamic change of self-organized network topology.When network size is large or change in topology is frequent, larger for being obtained from the expense that organization network topology information produces.

Summary of the invention

The object of the invention is to overcome the deficiency of above-mentioned prior art, the self-organizing network unicast method that a kind of network code based on cognitive combines with route is proposed, with business interval time of advent change at random in self-organizing network, transmission of radio links is made mistakes and network topology dynamic change in the situation that, reduce end-to-end time delay, effectively utilize network code in route advantage separately, reduce to obtain the expense that self-organized network topology produces, thereby improve the bearing capacity of self-organizing network to unicast service.

For achieving the above object, the method that the present invention proposes comprises following steps:

(1) each node in self-organizing network periodically sends parameter state frame to its neighbor node within the scope of double bounce around;

(2) node of receiving parameter state frame is according to the parameter information comprising in frame, calculate the neighbor node number within the scope of own double bounce, find Local network topology, and judge oneself whether to be included in butterfly topology according to the Local network topology of finding, if so, go to step (3); Otherwise, according to self-organizing network unicast method, process;

(3) node being included in butterfly topology judges whether oneself is key node, if so, and execution step (4); Otherwise execution step (5); Described key node refers to the node of two link infalls;

(4) the neighbor node parameter status information within the scope of the double bounce that the key node utilization in butterfly topology is obtained, calculates respectively the end-to-end time delay estimated value based on route clean culture strategy and network code clean culture strategy;

(4a) calculate based on route clean culture strategy end-to-end time delay estimated value

$D_{\mathrm{total}}^{\mathrm{Routing}}=4(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+4\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}$

$+4\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{(1-p_b)}^{l_D}}+{4T}_{\mathrm{CTS}}+{4T}_{\mathrm{SIFS}}-{4T}_{\mathrm{Timeout}}^{\mathrm{ACK}}$

In formula, p _rTSrepresent that node sends the probability of RTS grouping, these two calculation of parameter of the node traffic average arrival rate comprising in nodes n and parameter state frame obtain, and n represents nodes, CW _minexpression minimum is kept out of the way window, T _slotbe the length of a time slot, T _rTSthe propagation delay time that represents RTS grouping, T _sIFSrepresent short frame interFrameGap, T _aCKthe propagation delay time that represents ACK grouping, T _dIFSrepresent the interFrameGap in distributed coordination function, T _datathe propagation delay time that represents packet,

for the overtime thresholding of acknowledgement frame ACK, p _band l _drepresent respectively link bit error rate and block length, T _cTSthe propagation delay time that represents CTS grouping;

(4b) calculate coding clean culture strategy end-to-end time delay estimated value Network Based

$D_{\mathrm{total}}^{\mathrm{Coding}}=3(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+3\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}$

$+3\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{(1-p_b)}^{2l_D}}+{3T}_{\mathrm{CTS}}+{3T}_{\mathrm{SIFS}}-{3T}_{\mathrm{Timeout}}^{\mathrm{ACK}}$

In formula the implication of each symbol with

computing formula in identical;

(5) size of the more above-mentioned two kinds of clean culture strategy end-to-end time delay estimated values of key node, chooses the clean culture strategy that estimated value is lower, and by top node and bottom node in the clean culture strategy notice butterfly topology of choosing; When having grouping to arrive top node, key node, top node and bottom node adopt the clean culture strategy of choosing jointly to complete the transmission of grouping; During as transmission in grouping, there are abnormal conditions, to the processing of classifying of these abnormal conditions.

The present invention compares with existing scheme, and tool has the following advantages:

1. key node of the present invention is dynamically adjusted the clean culture strategy of choosing according to the variation of bit error rate, block length and the network topology of node traffic buty cycle, unlimited link, therefore has adaptivity.

2. key node of the present invention selects adaptively to have the clean culture strategy of lower end-to-end time delay estimated value between route clean culture strategy and network code clean culture strategy, therefore effectively utilized two kinds of clean culture strategies advantage separately, reduced to obtain the expense that self-organized network topology produces, improved the bearing capacity of self-organizing network to unicast service.

3. the present invention is all converted into by parameters such as business busy degree, link bit error rate, data packet length and interstitial contents the performance that end-to-end time delay is unified estimation network coding clean culture strategy and route clean culture strategy, is convenient to the two to quantize relatively.

Accompanying drawing explanation

Fig. 1 adopts in existing butterfly topology based on route clean culture strategy transmission schematic diagram data;

Fig. 2 adopts coding clean culture strategy transmission schematic diagram data Network Based in existing butterfly topology;

Fig. 3 is implementing procedure figure of the present invention;

Fig. 4 is the sub-process figure that in implementing procedure figure of the present invention, abnormal conditions classification is processed;

Fig. 5 is the present invention and existing route and the end-to-end time delay comparison diagram of network code clean culture strategy under different business buty cycle;

Fig. 6 be the present invention from existing route and network code clean culture strategy the end-to-end time delay comparison diagram under different bit error rates;

Fig. 7 the present invention and existing route and the end-to-end time delay comparison diagram of network code clean culture strategy under different grouping length;

Fig. 8 the present invention is the end-to-end time delay comparison diagram under different node densities from existing route and network code clean culture strategy.

Embodiment

Below by drawings and Examples, technical scheme of the present invention is described in further detail.

Application scenarios of the present invention is butterfly topology as depicted in figs. 1 and 2, wherein be positioned at the butterfly upper left node A of topology and top-right Node B and be called top node, the node C of link AC and BC infall is key node, and the node D and the bottom-right node E that are positioned at butterfly topology lower left are called bottom node.Top node A unicasts to bottom node E by grouping a, and top node B unicasts to bottom node D by grouping b.Because transmission range is limited, grouping a and b will pass through key node C.Obviously, key node C place can carry out network code.Here ignore operation time of key node C place coding and bottom node D, decode operation time in E place.

For simplifying the invention process step, make the following assumptions here:

1) the carrier sense scope of all nodes is double bounces, and transmission range is a jumping;

2) all groupings of transmitting in network are isometric;

3) re-transmission of grouping is all that error code by wireless link causes, and the bit error rate of every link is identical;

4) all nodes are operated in saturation condition, and all nodes continue to have grouping to send, and the average arrival rate of each node traffic is identical;

5) nodes density is that the neighbor node number within the scope of each node double bounce is all identical uniformly;

6) system works is under time slot state, and all nodes are all in synchronous regime.

For the butterfly topology shown in Fig. 1 and Fig. 2, the present invention to the cognitive unicast method process of transfer of data as shown in Figure 3.

With reference to Fig. 3, the process of cognitive unicast method that the present invention carries is as follows:

Step 1, Topology Discovery and parameter perception.

1.1) each node in self-organizing network periodically sends parameter state frame to its neighbor node within the scope of double bounce around, and wherein parameter state frame comprises: the average arrival rate of business, the bit error rate p of adjacent link _band data packet length l _d;

1.2) receive that the neighbor node number that the node of parameter state frame calculates within the scope of own double bounce is node density n, find the Local network topology within the scope of own double bounce, and judge in the butterfly topology oneself whether being included in as shown in Figure 1, if so, perform step 2; Otherwise process according to existing self-organizing network unicast method;

Step 2, the node being included in butterfly topology judges whether oneself is key node, and the node C in Fig. 1, if so, performs step 3; Otherwise, perform step 4.

Step 3, key node C calculates respectively the end-to-end time delay estimated value based on route clean culture strategy and coding clean culture strategy Network Based.

3.1) calculate the end-to-end time delay estimated value of route clean culture strategy

According to Fig. 1, build the end-to-end time delay estimated value of route clean culture strategy

computing formula:

As shown in Figure 1,

be the time delay estimated value sum of four shared channels of single-hop transmission:

$D_{\mathrm{total}}^{\mathrm{Routing}}=D_{A\_C}^{\mathrm{Routing}}+D_{C\_E}^{\mathrm{Routing}}+D_{B\_C}^{\mathrm{Routing}}+D_{C\_D}^{\mathrm{Routing}}$

In formula,

for top node A transfers to grouping a the time delay estimated value of the shared channel of key node C,

for key node C is forwarded to the grouping a receiving the time delay estimated value of the shared channel of bottom node E,

for top node B transfers to grouping b the time delay estimated value of the shared channel of key node C,

for top node B is forwarded to grouping b the time delay estimated value of the shared channel of key node C.

In self-organizing network, all nodes are all with CSMA/CA mechanism access channel, and the carrier sense of channel and grouping send all take time slot as base unit.According to RTS/CTS handshake mechanism in IEEE802.11DCF rule, before transmission data, sending node and receiving node first by RTS/CTS frame preengage alternately channel, the binary exponential backoff algorithm that RTS frame clashes in rear employing DCF rule is processed, and finally obtains

computing formula:

$D_{A\_C}^{\mathrm{Routing}}=(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}$

$+\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{(1-p_b)}^{l_D}}+T_{\mathrm{CTS}}+T_{\mathrm{SIFS}}-T_{\mathrm{Timeout}}^{\mathrm{ACK}}$

In formula, p _rTSrepresent that node A sends the probability of RTS grouping, be the business buty cycle of node A, the formula that these two parameter informations of the neighbor node number n within the scope of the node traffic average arrival rate comprising in the parameter state frame that can be sent by node A and the double bounce of node A provide according to document " Performance Analysis of the IEEE 802.11Distributed Coordination Function " calculates; N represents the neighbor node number n within the scope of the double bounce of node A, i.e. node density, and it is calculated according to the parameter state frame of receiving by node A; CW _minexpression minimum is kept out of the way window, T _slotbe the length of a time slot, T _rTSrepresent RTS frame propagation delay time, T _sIFSrepresent short frame interFrameGap, T _aCKthe propagation delay time that represents ACK frame, T _dIFSrepresent the interFrameGap in distributed coordination function, T _datathe propagation delay time that represents grouping,

for the overtime thresholding of acknowledgement frame ACK, above-mentioned parameter provides by IEEE802.11 agreement; p _band l _drepresent respectively the bit error rate of link AC and the block length that node A sends, the parameter state frame that can be sent by node A obtains; T _cTSthe propagation delay time that represents CTS frame, it is provided by IEEE 802.11 agreements.

According to hypothesis above, for node A, B, C, D and E, p _rTS, n and l _dall identical, for link AC, BC, CD and CE, p _ball identical, so

with

expression formula all identical, therefore have:

$D_{\mathrm{total}}^{\mathrm{Routing}}=D_{A\_C}^{\mathrm{Routing}}+D_{C\_E}^{\mathrm{Routing}}+D_{B\_C}^{\mathrm{Routing}}+D_{C\_D}^{\mathrm{Routing}}$

$={4D}_{A\_C}^{\mathrm{Routing}}$

$=4(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+4\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}$

$+4\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{(1-p_b)}^{l_D}}+{4T}_{\mathrm{CTS}}+{4T}_{\mathrm{SIFS}}-{4T}_{\mathrm{Timeout}}^{\mathrm{ACK}}$

3.2) the end-to-end time delay estimated value of computing network coding clean culture strategy

As shown in Figure 1,

be three the shared channel delay estimated value of single-hop transmission sums:

$D_{\mathrm{total}}^{\mathrm{Coding}}=D_{A\_\mathrm{CD}}^{\mathrm{Coding}}+D_{B\_\mathrm{CE}}^{\mathrm{Coding}}+D_{C\_\mathrm{DE}}^{\mathrm{Coding}}$

In formula,

represent that top node A transfers to grouping a the time delay estimated value of key node C and the shared channel of bottom node D; represent that top node B transfers to grouping b the time delay estimated value of key node C and the shared channel of bottom node E;

represent that key node C is by the grouping after coding

transfer to the time delay estimated value of bottom node D, the shared channel of E.

Equally, according to DCF rule, obtain computing formula:

$D_{A\_\mathrm{CD}}^{\mathrm{Coding}}=(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}$

$+\frac{T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}+T_{\mathrm{SIFS}}}{{(1-p_b)}^{2l_D}}+T_{\mathrm{CTS}}+T_{\mathrm{SIFS}}-T_{\mathrm{Timeout}}^{\mathrm{ACK}}$

In above formula, p _bboth the bit error rate that had represented link AC, AD, the implication of all the other symbols with

identical in expression formula.

Same foundation hypothesis above,

with

expression formula all identical, therefore have:

$D_{\mathrm{total}}^{\mathrm{Coding}}=D_{A\_\mathrm{CD}}^{\mathrm{Coding}}+D_{B\_\mathrm{CE}}^{\mathrm{Coding}}+D_{C\_\mathrm{DE}}^{\mathrm{Coding}}$

$={2D}_{A\_\mathrm{CD}}^{\mathrm{Coding}}$

$=3(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+3\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}.$

$+3\frac{T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}+T_{\mathrm{SIFS}}}{{(1-p_b)}^{2l_D}}+{3T}_{\mathrm{CTS}}+{3T}_{\mathrm{SIFS}}-{3T}_{\mathrm{Timeout}}^{\mathrm{ACK}}$

Step 4, clean culture strategic decision-making.

Key node C in butterfly topology is the end-to-end time delay estimated value of route clean culture strategy relatively

end-to-end time delay estimated value with network code clean culture strategy

choose the clean culture strategy that end-to-end time delay estimated value is lower, and by top node A, B and bottom node D, E in the clean culture strategy notice butterfly topology of choosing.

When key node C is when selecting to have the clean culture strategy of lower end-to-end time delay estimated value, there will be following two kinds of situations:

(1) when adoption Network coding clean culture strategy transmission grouping;

(2) when

adopt route clean culture strategy transmission grouping.

From step 4, key node C always chooses the clean culture strategy with lower end-to-end time delay estimated value and transmits grouping, so the end-to-end time delay estimated value D of institute of the present invention extracting method _totalfor:

$D_{\mathrm{total}}=\min \{D_{\mathrm{total}}^{\mathrm{Routing}},D_{\mathrm{total}}^{\mathrm{Routing}}\}$

Step 5, clean culture strategy execution.

When having grouping to arrive top node, key node C, top node A, B and bottom node D, E adopt the lower clean culture strategy of clean culture time delay estimated value of choosing jointly to complete the transmission of grouping.

5.1), if choose route clean culture strategy, after the grouping a (b) that key node C sends at the top node A (B) that receives butterfly topology, directly forward it to the bottom node E (D) of butterfly topology;

5.2), if choose network code clean culture strategy, key node C will wait until that the grouping a of top node A transmission and the grouping b that top node B sends arrive the network code mode of carrying out again XOR afterwards, and by the grouping after coding

transfer to two bottom node D, E.For assurance node D, E place, successfully decode, require node D must receive grouping a, node E must receive grouping b.Like this, bottom node D is by computing

obtain required grouping b, bottom node E is by computing

obtain required grouping a.

5.3), with reference to Fig. 4, when Adoption Network coding clean culture strategy, the following three kinds of abnormal conditions that occur are processed:

First kind situation, the grouping a (b) that top node A (B) sends arrives after key node C, and the grouping b (a) that another one top node B (A) sends is at the overtime thresholding of regulation

while inside not arriving key node C, key node C is directly forwarded to bottom node B (A) by a grouping a (b) who receives; Overtime thresholding wherein

after referring to grouping a (b) the arrival key node C of a butterfly top node A of topology (B) transmission, the grouping b (a) that another top node B (A) sends arrives the time that key node C needs, therefore with

implication by step 3, provided;

Equations of The Second Kind situation, has the decoding that fails at least in two bottom node D, E in butterfly topology, coding clean culture strategy exits network;

The 3rd class situation, has node to leave in butterfly topology or new node adds.

Effect of the present invention can further illustrate by following simulation result.

1. simulating scenes: Fig. 1 and Fig. 2 are the simulating scenes of the inventive method.Wherein Fig. 5, Fig. 6 and Fig. 7 are all at node density n=4, the topology that is self-organizing network is to only have under the situation of butterfly topology of five nodes the end-to-end time delay of the present invention and existing route and network code clean culture strategy to carry out simulation comparison, and Fig. 8 carries out simulation comparison by the end-to-end time delay of the present invention and existing route and network code clean culture strategy under the situation evenly increasing at node density.

2. simulated conditions: emulation is carried out according to RTS/CTS handshake mechanism in IEEE802.11DCF pattern.In IEEE802.11 agreement, Frame is different from the transmission rate of control frame, take 802.11a as example, T _slot=9 μ s, T _sIFS=16 μ s, the packet of employing and control packet RTS, CTS transmission rate are respectively 24Mbit/s and 9Mbit/s, and the propagation delay time of control packet RTS, CTS and acknowledgment packet ACK and packet can be obtained by IEEE802.11 agreement.

3. emulation content and simulation result:

Emulation one: by the present invention and existing route and network code clean culture strategy at different business buty cycle p _rTSunder end-to-end time delay contrast, simulation result is as shown in Figure 5; Fig. 5 shows, works as p _rTSduring < 0.07, when node traffic is not busy,

be applicable to adopting route clean culture strategy.

Emulation two: by the present invention and existing route and network code clean culture strategy at different bit error rate p _bunder end-to-end time delay contrast, simulation result is as shown in Figure 6; Fig. 6 shows, works as p _b< 5 * 10 ^-5time, when link reliability is higher, being applicable to Adoption Network encoding scheme transmission data, visible network coding is higher to the reliability requirement of link.

Emulation three: by the present invention and existing route and network code clean culture strategy at different pieces of information block length l _dunder end-to-end time delay contrast, simulation result is as shown in Figure 7; Fig. 7 shows, works as l _dduring < 6000bit,

visible network coding is applicable to the shorter grouping of transmission.

Emulation four: the end-to-end time delay under different node density n contrasts from existing route and network code clean culture strategy by the present invention, and simulation result as shown in Figure 8; Fig. 8 shows, when n > 7,

visible network coding is applicable in network that node density is larger.

Above-mentioned four emulation show: the method that the present invention proposes is to adjust adaptively clean culture strategy according to the dynamic change of node traffic, wireless losses rate and network topology, effectively utilize two kinds of clean culture strategies of network code and route advantage separately, improved the bearing capacity of self-organizing network to unicast service; Minimum between the end-to-end time delay estimated value always route clean culture strategy of the method and network code clean culture strategy, optimized the end-to-end time delay of network, improved network performance.

Term explanation

CSMA/CA:Carrier Sense Multiple Access/Collision Avoidance, the carrier sense multiple access protocol that has conflict to avoid;

IEEE:Institute of Electrical and Electronics Engineers, IEEE-USA;

DCF:Distributed Coordination Function, distributed coordination function;

RTS:Request To Send, request sends frame;

CTS:Clear To Send, allows to send frame;

ACK:Acknowledgement Character, affirmative acknowledgement.

Claims (6)

1. the self-organizing network unicast method that the network code based on cognitive combines with route, comprises following steps:

(1) suppose following condition:

The carrier sense scope of all nodes is double bounces, and transmission range is a jumping; All groupings of transmitting in network are isometric; The re-transmission of grouping is all that the error code by wireless link causes, and the bit error rate of every link is identical; All nodes are operated in saturation condition, and all nodes continue to have grouping to send, and the average arrival rate of each node traffic is identical; Nodes density is that the neighbor node number within the scope of each node double bounce is all identical uniformly; System works is under time slot state, and all nodes are all in synchronous regime;

(2) each node in self-organizing network periodically sends parameter state frame to its neighbor node within the scope of double bounce around;

(3) node of receiving parameter state frame is according to the parameter information comprising in frame, calculate the neighbor node number within the scope of own double bounce, find Local network topology, and judge oneself whether to be included in butterfly topology according to the Local network topology of finding, if so, go to step (4); Otherwise, according to self-organizing network unicast method, process;

(4) node being included in butterfly topology judges whether oneself is key node, if so, and execution step (5); Otherwise execution step (7); Described key node refers to the node of two link infalls;

(5) the neighbor node parameter status information within the scope of the double bounce that the key node utilization in butterfly topology is obtained, calculates respectively the end-to-end time delay estimated value based on route clean culture strategy and network code clean culture strategy;

(5a) calculate based on route clean culture strategy end-to-end time delay estimated value

$\begin{array}{c}{D}_{\mathrm{total}}^{\mathrm{Routing}}=4(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+4\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}\\ +4\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{\left({1-p}_b\right)}^{l_D}}+{4T}_{\mathrm{CTS}}+{4T}_{\mathrm{SIFS}}-{4T}_{\mathrm{Timeout}}^{\mathrm{ACK}}\end{array}$

In formula, p _rTSrepresent that node sends the probability of RTS grouping, these two calculation of parameter of the node traffic average arrival rate comprising in nodes n and parameter state frame obtain, and n represents the neighbor node number within the scope of the double bounce of each node, CW _minexpression minimum is kept out of the way window, T _slotbe the length of a time slot, T _rTSthe propagation delay time that represents RTS grouping, T _sIFSrepresent short frame interFrameGap, T _aCKthe propagation delay time that represents ACK grouping, T _dIFSrepresent the interFrameGap in distributed coordination function, T _datathe propagation delay time that represents packet,

for the overtime thresholding of acknowledgement frame ACK, p _band l _drepresent respectively link bit error rate and block length, T _cTSthe propagation delay time that represents CTS grouping;

(5b) calculate coding clean culture strategy end-to-end time delay estimated value Network Based

$\begin{array}{c}{D}_{\mathrm{total}}^{\mathrm{Coding}}=3(2^{\frac{1}{{(1-p_{\mathrm{RTS}})}^n}}-1){\mathrm{CW}}_{\min }{T}_{\mathrm{slot}}+3\frac{T_{\mathrm{RTS}}+T_{\mathrm{SIFS}}+T_{\mathrm{ACK}}+T_{\mathrm{DIFS}}}{{(1-p_{\mathrm{RTS}})}^n}\\ +3\frac{T_{\mathrm{SIFS}}+T_{\mathrm{Data}}+T_{\mathrm{Timeout}}^{\mathrm{ACK}}}{{\left({1-p}_b\right)}^{2l_D}}+{3T}_{\mathrm{CTS}}+{3T}_{\mathrm{SIFS}}-{3T}_{\mathrm{Timeout}}^{\mathrm{ACK}}\end{array}$

In formula the implication of each symbol with

computing formula in identical;

(6) size of the more above-mentioned two kinds of clean culture strategy end-to-end time delay estimated values of key node, chooses the clean culture strategy that estimated value is lower, and by top node and bottom node in the clean culture strategy notice butterfly topology of choosing;

(7), when having grouping to arrive top node, key node, top node and bottom node adopt the clean culture strategy of choosing jointly to complete the transmission of grouping; During as transmission in grouping, there are abnormal conditions, to the processing of classifying of these abnormal conditions.

2. the self-organizing network unicast method that the network code based on cognitive according to claim 1 combines with route, wherein the related parameter state frame of step (2) comprises: the average arrival rate of business, the bit error rate of adjacent link and data packet length.

3. the self-organizing network unicast method that the network code based on cognitive according to claim 1 combines with route, wherein the related top node of step (6), refers to and lays respectively at butterfly topology upper left side and top-right two nodes.

4. the self-organizing network unicast method that the network code based on cognitive according to claim 1 combines with route, wherein the related bottom node of step (6), refers to and lays respectively at butterfly topology lower left and bottom-right two nodes.

5. the self-organizing network unicast method that the network code based on cognitive according to claim 1 combines with route, wherein related abnormal conditions and the classification processing of step (7) comprises following three classes:

First kind situation, key node has been chosen the clean culture strategy of coding Network Based, but after the grouping arrival key node that top node of butterfly topology sends, while not arriving key node in the overtime thresholding that is grouped in regulation that another one top node sends, key node directly by a forwarding of packets of receiving to bottom node;

Equations of The Second Kind situation, key node has been chosen the clean culture strategy of coding Network Based, but has the decoding that fails at least in two bottom node of butterfly topology, and coding clean culture strategy exits network;

The 3rd class situation, key node has been chosen the clean culture strategy of coding Network Based, but in butterfly topology, has node to leave or new node adds, and coding clean culture strategy exits network.

6. the self-organizing network unicast method that the network code based on cognitive according to claim 5 combines with route, the overtime thresholding of related regulation in first kind situation wherein, refer to that the grouping that another top node sends arrives the time that key node needs after the grouping arrival key node that top node of butterfly topology sends.

Images