CN102710448A - Adaptive regulating method of life cycle parameter of SAVI (Source Address Validation Improvements) switch filter table - Google Patents

Abstract

The invention discloses an adaptive regulating method of a filter table life cycle parameter of an SAVI (Source Address Validation Improvements) switch. The adaptive regulating method comprises the steps of: setting a filter table life cycle parameter of a switch, collecting a user online record of the switch and a load state of a filter table of the switch within a preset work cycle; obtaining a plurality of standby filter table life cycle parameter values; selecting an optimal value of an initial filter table life cycle parameter and setting a filter table item number regulating threshold; in a first sampling point of each work life cycle of the operation of the switch, based on the operation condition of the filter table in the last work life cycle, calculating the optimal value of the filter table life cycle parameter in the work life cycle; and in each sampling point, calculating an average table item number of the filter table and comparing with a filter table item number regulating threshold, and regulating the filter table life cycle parameter in real time. The invention can be used for dynamically regulating the filter table life cycle parameter.

Description

SAVI switch filter table parameter adaptive adjusting method life cycle

Technical field

The present invention relates to technical field of computer network management, particularly a kind of SAVI switch filter table parameter adaptive adjusting method life cycle.

Background technology

SAVI (Source Address Validation Improvements, source address proof scheme) provides a kind of host-level other source address proof scheme, and this scheme is deployed on the access device.The core concept that the SAVI switch carries out the source address checking is through monitoring the procotol message relevant with the IPv6 address assignment; With the MAC Address of the IPv6 address of main frame through legal means configuration, this network interface card and on connect port numbers and deposit in the binding table of switch, and download in the filter table of switch.Whenever under when connecting port and having message to pass through, this filter table of exchange opportunity inquiry is to confirm whether this message can pass through.

These filter table all are to use hardware such as TCAM, because the hardware resource price comparison is expensive, the price of equipment such as while access-layer switch again can not be too high, so the quantity of filtering meter item is very limited usually, can only support a hundreds of filtering meter item usually.Exchange opportunity is that each list item in the filter table is provided with life cycle, and when the existence expiration, whether also exchange opportunity is sent NS request (Neighbor Solicitation, neighbor request) and judged the survival of the corresponding main frame of this list item.

The deployment way of SAVI switch is divided into two kinds of cascade and non-cascades, and is as shown in Figure 1.The cascade deployment way is to connect other general switch under the SAVI switch, and shown in Fig. 1 (a), but not the cascade deployment way is directly to connect main frame under the SAVI switch, shown in Fig. 1 (b).For the deployment scenario of non-cascade, switch can detect the down incident of access interface, thus the deletion of filtering meter item that rapidly will be relevant with this port.Yet this mode is under the situation that cascade is disposed and infeasible, because what connect under the SAVI switch this moment is general switch, and this switch can not cut off the power supply usually, thereby the SAVI switch can't detect the down incident of port.Therefore after having only the existence expiration of waiting for filtering meter item, when switch is confirmed process to this list item, can find that just the corresponding main frame of this list item leaves, thereby with this list item deletion.

Therefore, under the cascade deployment way, it is very important that filtering meter item parameter life cycle is provided with a rational value.If what should value be provided with is too little, each filtering meter item expiration back switch of at every turn surviving all will carry out the affirmation process so, and this will increase the load of switch, and the while also can consume certain network bandwidth; And if should value be provided with too big, then switch just can not detect leaving of user timely, therefore useless filtering meter item will take the filter table resource of preciousness.Worse, if filter table is filled, so new filtering meter item can't be received in, and cause legal message also to be dropped, thereby cause the serious consequence of network denial of service.

Summary of the invention

The object of the invention is intended to solve at least one of above-mentioned technological deficiency, proposes a kind of SAVI switch filter table parameter adaptive adjusting method life cycle especially, and this method can dynamically be adjusted filter table parameter life cycle.

For achieving the above object, embodiments of the invention have proposed a kind of SAVI switch filter table parameter adaptive adjusting method life cycle, comprise the steps:

Filter table parameter life cycle of SAVI switch is set, in said filter table parameter life cycle, collects the load condition of filter table of user's online record and the said SAVI switch of said SAVI switch in preset time work period;

According to the user's online record of said SAVI switch and the load condition of filter table, obtain a plurality of alternative filter table parameter values life cycle;

From said a plurality of alternative filter table parameter values life cycle, choose the inceptive filtering hypergene and deposit the period parameters optimal value, and filter table item number adjustment threshold value is set;

First sampled point in each said time work period that said SAVI switch moves based on the ruuning situation of a last work period temporal filtering table, calculates the filter table parameter life cycle optimal value of time this work period; And

Each sampled point in said time work period; Calculate the average list item number of said filter table, and the average list item number and the filter table item number adjustment threshold value of said filter table compared, and adjust said filter table parameter life cycle in real time according to comparative result; Wherein, At current sampling point, when the average list item number of said filter table is adjusted threshold value greater than said filter table item number, current filter table parameter value life cycle is turned down rapidly; Otherwise; If at current sampling point; When the average list item number of said filter table is adjusted threshold value less than said filter table item number; And filter table parameter life cycle of a last sampled point is during less than the optimal value of said filter table parameter life cycle, current filter table parameter value life cycle is transferred big, but this value does not allow the optimal value above said filter table parameter life cycle.

SAVI switch filter table parameter adaptive adjusting method life cycle according to the embodiment of the invention; Can be under the deployment way of SAVI switch in non-cascade; Variation according to filter table item number in the SAVI switch is dynamically adjusted filtering meter item parameter life cycle, is guaranteeing to reduce the number of duplicate acknowledgments of switch to filtering meter item as far as possible under the prerequisite that the switch filter table can not be filled at any time.

Aspect that the present invention adds and advantage part in the following description provide, and part will become obviously from the following description, or recognize through practice of the present invention.

Description of drawings

Above-mentioned and/or additional aspect of the present invention and advantage are from obviously with easily understanding becoming the description of embodiment below in conjunction with accompanying drawing, wherein:

Fig. 1 (a) is the sketch map of the non-cascade deployment way of SAVI switch;

Fig. 1 (b) is the sketch map of the cascade deployment way of SAVI switch;

Fig. 2 is the flow chart of SAVI switch filter table parameter adaptive adjusting method life cycle of the embodiment of the invention;

Fig. 3 is the CDF curve of filter table item number under the various configuring conditions of the embodiment of the invention;

Fig. 4 is Avg under the various configuring conditions of the embodiment of the invention _DCThe CDF curve;

Fig. 5 is R under the various configuring conditions of the embodiment of the invention _WThe CDF curve;

Fig. 6 is the SAVI switch three all inner filtration list item numbers and the parameter value life cycle situation over time of the high capacity of the embodiment of the invention;

Fig. 7 is the SAVI switch three all inner filtration list item numbers and the parameter value life cycle situation over time of the low load of the embodiment of the invention;

Fig. 8 is that the SAVI switch of high capacity of the embodiment of the invention is in a burst time period inner filtration list item number and parameter value life cycle situation over time; And

Fig. 9 is that the SAVI switch of low load of the embodiment of the invention is in a burst time period inner filtration list item number and parameter value life cycle situation over time.

Embodiment

Describe embodiments of the invention below in detail, the example of said embodiment is shown in the drawings, and wherein identical from start to finish or similar label is represented identical or similar elements or the element with identical or similar functions.Be exemplary through the embodiment that is described with reference to the drawings below, only be used to explain the present invention, and can not be interpreted as limitation of the present invention.

Disclosing of hereinafter provides many various embodiment or example to be used for realizing different structure of the present invention.Of the present invention open in order to simplify, hereinafter the parts and the setting of specific examples are described.Certainly, they only are example, and purpose does not lie in restriction the present invention.In addition, the present invention can be in different examples repeat reference numerals and/or letter.This repetition is in order to simplify and purpose clearly, itself not indicate the relation between various embodiment that discuss of institute and/or the setting.In addition, various specific technology and the examples of material that the invention provides, but those of ordinary skills can recognize the property of can be applicable to of other technologies and/or the use of other materials.In addition; First characteristic of below describing second characteristic it " on " structure can comprise that first and second characteristics form the embodiment of direct contact; Can comprise that also additional features is formed on the embodiment between first and second characteristics, such first and second characteristics possibly not be direct contacts.

In description of the invention, need to prove, unless otherwise prescribed and limit; Term " installation ", " linking to each other ", " connection " should be done broad understanding, for example, can be mechanical connection or electrical connection; Also can be the connection of two element internals, can be directly to link to each other, and also can link to each other indirectly through intermediary; For those of ordinary skill in the art, can understand the concrete implication of above-mentioned term as the case may be.

With reference to following description and accompanying drawing, with these and other aspects of knowing embodiments of the invention.These describe with accompanying drawing in, some specific implementations in the embodiments of the invention are specifically disclosed, represent some modes of principle of the embodiment of embodiment of the present invention, still should be appreciated that the scope of embodiments of the invention is not limited.On the contrary, embodiments of the invention comprise and fall into appended spirit that adds the right claim and all changes, modification and the equivalent in the intension scope.

As shown in Figure 2, SAVI switch filter table parameter adaptive adjusting method life cycle according to an embodiment of the invention may further comprise the steps:

S101: filter table parameter life cycle of SAVI switch is set, in filter table parameter life cycle, collects the load condition of filter table of user's online record and the SAVI switch of SAVI switch in preset time work period

In embodiments of the present invention, time work period is set to a week, promptly 7 days.It will be appreciated that the time work period example in an above-mentioned week is just for the ease of describing the embodiment of the invention, and should not be construed as limitation of the present invention.

At first following several notions are described:

(1) filter table exhausting rate (R _O): being located at has N in a period of time _AThe individual sampling period is if wherein there is N _OThe resource of individual sampling period filter table is exhausted, then definition

(2) the useless resources occupation rate (R of filter table _W): be located in a period of time, all filtering meter item quilts are from 1 to n mark.If a filtering meter item i is added in the filter table at

time point; Deleted at

time point; And the main frame related with this filtering meter item then defines at

time point deviated from network

$S_W={\mathrm{\Σ}}_{i=1}^n(t_{2_i}-t_{3_i}),$ $S_A={\mathrm{\Σ}}_{i=1}^n(t_{2_i}-t_{1_i}),$ $R_W=\frac{S_W}{S_A}.$

(3) average number of duplicate acknowledgments (Avg _DC): in a period of time, all filtering meter item quilts are from 1 to n mark.If a filtering meter item existence expiration, then the SAVI exchange opportunity is sent NS and is asked the main frame to correspondence.If this source address still is used, then main frame is used NA (neighbor advertisement) message response.After receiving response message, exchange opportunity is according to the life cycle of this list item of parameter value replacement life cycle of current configuration.For example; If parameter value life cycle of configuration is 5 minutes, and the line duration of a main frame is 17 minutes, switch is met the affirmation process of carrying out 4 times together to this list item one so; Wherein for once be necessary under ecotopia, therefore the number of duplicate acknowledgments to this list item is 3 times.Number of duplicate acknowledgments to list item i is designated as

And average number of duplicate acknowledgments is defined as ${\mathrm{Avg}}_{\mathrm{DC}}=\frac{{\mathrm{\Σ}}_{i=1}^n{N}_{\mathrm{DC}}^i}{n}.$

Along with the increase of filter table parameter value life cycle that disposes, although Avg _DCCan reduce, but R _WCan increase, simultaneously R _O>0 possibility also can increase.R _O>0 means that network took place owing to filter table is filled the new user unaccepted situation that networks that causes in certain time period.

In certain period of SAVI switch operation, if R _O>0, R _WVery big, and Avg _DCVery little, can draw: in some time period, filter table is filled, and the situation of denial of service probably took place, yet a lot of expired useless list item has taken limited filter table resource.Therefore need adjustment configuration parameter life cycle, it is become smaller value.

In certain period of SAVI switch operation, if R _O>0, R _WVery little, and Avg _DCVery big, can draw: in some time period, filter table is filled, but because Avg _DCVery big and R _WVery little, therefore selecting a smaller value for filter table configuration parameter life cycle will be very little to the burden contribution that alleviates filter table.This situation mainly is owing to connecting main frame has used too many filtering meter item down.In order to find this abnormal conditions, whether the user record of utilizing the operation monitoring module analysis to collect finds to exist a plurality of filtering meter items that same MAC Address is all arranged.If existence, then this operation monitoring module can be adjusted the filter table item number upper limit that the port that connects this abnormal host allows generation automatically; If do not exist, then this switch operates under the situation of high capacity really, therefore gives the alarm, and the new SAVI switch of suggestion administrator deploys comes load sharing.

In certain period of SAVI switch operation, if R _O=0, Avg _DCVery little, can draw: this SAVI switch operates under the good state.

To find the solution every switch initial optimum life cycle of parameter value and can be modeled as following optimization problem, suc as formula 1 to formula 3, though this optimization problem is not directly with R _WAs a restrictive condition, but because R _WWith R _OClosely related, to R _ORetrain, be equivalent to indirectly to R _WRetrain, therefore here for problem is simplified, only with R _OAs restrictive condition.

Min?Avg _DC(t) (1)

s.t.R _O(t)=0 (2)

Min _lifeTime≤t≤Max _lifeTime (3)

Avg _DC(t) and R _O(t) represent index Avg when life cycle, configuration parameter value was set as t respectively _DCAnd R _OValue, and from the consideration on fail safe and the performance, t can not surpass Max _Lifetime, can not be lower than Min _Lifetime, and Max _LifetimeAnd Min _LifetimeValue then set by the network manager.

Along with R _WThe increase of value, R _O>0 possibility also can increase.In order to guarantee R _O=0, when selecting optimum life cycle during configuration parameter value, R _WValue need be as far as possible little.Therefore further convert above-mentioned optimization problem to following Multiobjective Optimization Problem, suc as formula 4 to formula 6.Promptly obtain a plurality of alternative filter table parameter values life cycle.Same R _W(t) representative index R when life cycle, configuration parameter value was set as t _WValue.

Min?Avg _DC(t) (4)

Min?R _W(t) (5)

s.t.Min _lifeTime≤t≤Max _lifeTime (6)

Collect in the week user's online record more accurately through filter table configuration parameter life cycle being arranged to a very little value; Be reaching the standard grade and rolling off the production line the time of each user; With and the place of reaching the standard grade etc., user's online record can draw through collection and the analysis to list item in the filter table.Because filter table configuration parameter life cycle is arranged to smaller value; Therefore the SAVI switch also can be deleted expired useless list item in the filter table as early as possible under the situation that cascade is disposed, so user's online record that the SAVI network management system collects is similar to the accurate description actual conditions.Can find out initial optimum configuration parameter value life cycle of every SAVI switch based on user's online record in this week then.

In one embodiment of the invention, filter table parameter life cycle that the SAVI switch is set is 5 minutes.

S102:, obtain the set of alternative filter table parameter value life cycle according to the user's online record of SAVI switch and the load condition of filter table.

In order to address the above problem, for the filter table item number is provided with an adjustment threshold value, this value can actual loading be provided with according to network by the keeper in this step.In one embodiment of the invention, filter table item number adjustment threshold value is set, comprises: the weight that influences filtering meter item mean value, the total capacity of filter table, maximum, the sampling period that life cycle, parameter allowed are set, and the first threshold number of times.

Obtain a plurality of alternative filter table parameter values life cycle and be divided into two kinds of situation:

First kind of situation: for being less than or equal to Max _Lifetime, and more than or equal to Min _Lifetime, and can be by sampling period P _SamplingThe all values of dividing exactly, respectively as parameter value life cycle of configuration, and for above-mentioned each of parameter value, handle as follows: at first the average list item number of the filter table number of times that exceeds the quata is set to 0 life cycle.Each sampled point in preset time work period; Calculate its average filtration list item number, and should be worth with filter table item number adjustment threshold ratio, if average filtration list item number surpasses threshold value; Then the average list item number of the filter table number of times that exceeds the quata is added 1; If after each sampled point calculating in preset time work period, the average list item number of filter table exceeds the quata number of times less than the first threshold number of times, then it is joined in the set of alternative filter table parameter value life cycle.

Second kind of situation: if for first kind of situation, the set of alternative filter table parameter value life cycle that calculates is then explained for being less than or equal to Max for empty _Lifetime, and more than or equal to Min _Lifetime, and can be by sampling period P _SamplingThe all values of dividing exactly all can not satisfy at the average list item number of preset time work period inner filtration table and exceeds the quata number of times less than the first threshold number of times, and will be less than or equal to Max this moment earlier _Lifetime, and more than or equal to Min _Lifetime, and can be by sampling period P _SamplingThe all values of dividing exactly all joins in the set of alternative filter table parameter value life cycle.

S103: from the set of alternative filter table parameter value life cycle, choose the inceptive filtering hypergene and deposit the period parameters optimal value.Be elaborated in the face of calculating initial optimum filter table parameter life cycle step down.

Initial optimum filter table parameter value life cycle of same calculating also is divided into two kinds of situation, corresponds respectively to two kinds of situation among the S102.

First kind of situation: for each element in the set of alternative filter table parameter value life cycle, with it as behind the life cycle parameter value, for the average number of duplicate acknowledgments (Avg that in preset time work period, calculates this switch _DC), select Avg _DCMinimum value is as initial optimum filter table parameter value life cycle.If the corresponding Avg of a plurality of values is arranged _DCEquate, then calculate the corresponding useless resources occupation rate (R of filter table of these values again _W), from these values, select R _WMinimum value is as initial optimum filter table parameter value life cycle.

Second kind of situation: for each element in the set of alternative filter table parameter value life cycle, with it as behind the life cycle parameter value, for the useless resources occupation rate (R of filter table that in preset time work period, calculates this switch _W), select R _WMinimum value is as initial optimum filter table parameter value life cycle.If the corresponding R of a plurality of values is arranged _WEquate, then calculate the corresponding average number of duplicate acknowledgments (Avg of these values again _DC), from these values, select Avg _DCMinimum value is as initial optimum filter table parameter value life cycle.

In an embodiment of the present invention, the implementation of S101 ~ S103 step is deployed in the SAVI switch network management system.

After going out initial optimum filter table parameter life cycle to every SAVI switch computes, this value will be configured to every SAVI switch.

S104: first sampled point in each said time work period that the SAVI switch moves, based on the ruuning situation of a last work period temporal filtering table, calculate the filter table parameter life cycle optimal value of time this work period.

When said SAVI switch in last time work period; The average list item number of filter table exceeds the quata number of times less than preset said first threshold number of times; And during the maximum that the optimal value of said filter table parameter life cycle allows less than said filter table parameter life cycle in last time work period; The optimal value of last work period filter table parameter life cycle is added a time in sampling period, as the filter table parameter life cycle optimal value in work at present cycle.

When said SAVI switch in last time work period; The average list item number of filter table exceeds the quata number of times more than or equal to preset said first threshold number of times; And during the minimum value that the optimal value of said filter table parameter life cycle allows greater than said filter table parameter life cycle in last time work period; The optimal value of last work period filter table parameter life cycle is deducted a time in sampling period, as the filter table parameter life cycle optimal value in work at present cycle.

For dynamic adjustment phase, each switch filter table parameter life cycle optimal value has only one in time work period, be equivalent to a benchmark.Actual life cycle of the parameter value of most of time filter table is this filter table parameter life cycle optimal value.When emergency case occurring; Actual life cycle, parameter value just can be adjusted; After the burst in the past, this value also can be adjusted back " optimal value " gradually, and in order to guarantee that " optimal value " this benchmark can adjust along with actual motion; So, finely tune according to the practical operation situation in the last one-period at first sampled point of each time work period.

S105: at each sampled point of time work period, calculate the average list item number of filter table, and the average list item number and the filter table item number adjustment threshold value of filter table compared, and adjust filter table parameter life cycle in real time according to comparative result.

Particularly, through the current filter table item number of monitoring SAVI switch, calculate the average list item number of filter table, and formula (7) is seen in the definition of the average list item number of filter table, wherein avg _sThe average list item number of the filter table when being illustrated in sampled point S; Avg _S-1The average list item number of the filter table of calculating when being illustrated in sampled point S-1;

Expression filter table parameter configuration life cycle becomes lifetime, the filter table item number when sampled point s; w _nBe percentage weights, specify, be used for regulating historical data (avg by the keeper _S-1) calculating avg _sThe time influence.For fear of the utmost point sudden in the short time, when the average list item that calculates filter table is counted, adopted and asked weighted-average method with historical data.

${\mathrm{avg}}_s=(1-w_n)\·{\mathrm{avg}}_{s-1}+w_n\·N_s^{\mathrm{lifetime}}---\left(7\right)$

If the average list item number of filter table has surpassed the threshold value of administrator configurations, then current parameter life cycle is adjusted to less value, and the computational methods of concrete this value are seen formula (8), wherein avg _sBe the average list item number that utilizes the filter table that formula (7) calculates when sampled point s, lifetime _sFilter table parameter value life cycle that should use when being illustrated in sampled point s, lifetime _S-1Already used filter table parameter value life cycle when being illustrated in sampled point s-1; Otherwise add time in a sampling period (p for current parameter value life cycle _Sampling), this time in sampling period is meant the cycle of proxy polling filter table item number, i.e. lifetime _s=lifetime _S-1+ p _SamplingOptimal value then is the upper limit of parameter value life cycle.

${\mathrm{lifetime}}_s=\frac{c-{\mathrm{avg}}_s}{c}\·\mathrm{lifetim}{e}_{s-1}---\left(8\right)$

Wherein, lifetime _sBe filter table parameter life cycle of current sampling point, C is the total capacity of filter table, avg _sBe the current average list item number of filter table, lifetime _S-1Filter table parameter life cycle for a last sampled point.

At current sampling point; When the average list item number of filter table is adjusted threshold value less than the filter table item number; And filter table parameter life cycle of a last sampled point is during less than the optimal value of said filter table parameter life cycle; Filter table parameter life cycle of a last sampled point is added the time in sampling period, be set to filter table parameter life cycle of current sampling point.

In one embodiment of the invention, step S104 and step S105 realize in the agency of every SAVI switch.

Deposit choosing of period parameters optimal value in the face of the inceptive filtering hypergene down and the detailed step of the real-time adjustment of filter table parameter life cycle is described.

1, chooses the inceptive filtering hypergene and deposit period parameters

In the SAVI network management system, all carry out following set-up procedure for every SAVI switch:

Step (1): the keeper is provided with following parameter:

w _n: the weight that influences filtering meter item mean value;

C: the total capacity of filter table;

β: adjustment threshold value (percentage of total capacity);

Max _Lfetime:: the desirable maximum of configuration parameter life cycle;

Min _Lifetime:: the desirable minimum value of configuration parameter life cycle;

p _Sampling: the sampling period (minute to be unit);

Total: total the number of minutes in a week;

Count _Max: for certain of parameter value, this switch allows the average list item number of filter table to surpass the number of times of threshold value beta C in the week life cycle.

Step (2): the following parameter of initialization:

Set _Time: the set of optional filter table lifetime values, the element in this set is less than Max _Lifetime, greater than Min _Lifetime, simultaneously can be by p _SamplingDivide exactly.

Set _Sampling: the time of all sampled points set in a week, each element is minute to be unit, promptly each element simultaneously can be by p all less than Total _SamplingDivide exactly.

Candidate _Lifetime: the intermediate object program set when algorithm is carried out is initially sky.

Step (3): for Set _TimeEach element t in the set, carry out following steps successively:

Step (3.1): the count variable is set as 0, and this variable is used for expression for element t, and the average filtration list item number of this switch in this week is more than or equal to the number of times of threshold value beta C.

Step (3.2): for Set _SamplingEach element s in the set, carry out following steps successively:

Step (3.2.1): if s equals 0, promptly when first sampled point, the average list item of filter table is counted avg and is equaled

Wherein

If expression filter table parameter configuration life cycle becomes t, the filter table item number when sampled point s; If s is not equal to 0, promptly not when first sampled point, avg equals following result: 1 deducts w _nAfter the avg that calculates when multiply by a sampled point, add w at last _nMultiply by

Step (3.2.2): if the avg that calculates in the step (3.2.1) then adds 1 with the count variable greater than threshold value beta C.

Step (3.2.3): if count is greater than or equal to count _MaxIf, then the explanation with life cycle parameter configuration become t, the probability that filter table is filled is very big, so end step 3.2.

Step (3.3): if through after the step (3.2), the count variable is still less than count _Max, explain then that from this all user's online record select t as life cycle during parameter value, the probability that filter table is filled is less, can therefore t be joined Candidate with t as a chosen candidate value _LifetimeIn the set.

Step (4): if Candidate _LifetimeSet is not empty, explains through after the step (3), according to user's online record in a week, has selected a series of alternative parameter values life cycle, and when using these values to be configured, the probability that filter table is filled is all less.

From these a series of chosen candidate values, select an optimum value, therefore carry out following steps successively:

Step (4.1):

is set as infinity;

expression is when life cycle, parameter value was configured to optimal, in the average number of duplicate acknowledgments of all inner filtration list items.

Step (4.2): for Candidate _LifetimeEach element t in the set, carry out following steps successively:

Step (4.2.1): based on the user's online record in the week; Calculate

and represent when life cycle, parameter value was configured to t, in the average number of duplicate acknowledgments of all inner filtration list items.

Step (4.2.2): if less than

explanation when choosing t as life cycle during parameter value; Average number of duplicate acknowledgments when choosing optimal is little; Therefore optimal is set as t again, simultaneously

is set as again

Step (4.2.3): if

Equal

Then explanation is when choosing t and optimal as life cycle during parameter value, from average number of duplicate acknowledgments (Avg _DC) this index, the two is of equal value, considers the useless resources occupation rate (R of filter table this moment _W) this filter table operating index.Therefore carry out following steps successively:

Step (4.2.3.1): based on the user's online record in the week; Calculate respectively

and wherein

and

represent when life cycle, parameter value was configured to t and optimal, respectively in the useless resources occupation rate of all inner filtration tables.

Step (4.2.3.2): if

Less than

Then description selection t is as life cycle during parameter value, from the useless resources occupation rate (R of filter table _W) this index is better than and chooses optimal.Therefore optimal is set as t again.

Step (5): if Candidate _LifetimeSet, according to user's online record in a week, is found from Set after the then explanation process step (3) for empty _TimeChoose any one element in the set as parameter value life cycle; All can not guarantee at any one sampled point; The average filtration list item is counted avg less than threshold value beta C; It is higher as the probability that life cycle, parameter value all might cause being filled in certain sampled point filter table to be any one value, and this situation possibly be because more the causing of number of users that switch inserts down.Can only choose a value this moment, makes R _WAs much as possible little, promptly improve the service efficiency of filter table as far as possible.Therefore carry out following steps successively:

Step (5.1):

is set as infinity;

expression is when life cycle, parameter value was configured to optimal, in the useless resources occupation rate of all inner filtration tables.

Step (5.2): for Set _TimeEach element t in the set, carry out following steps successively:

Step (5.2.1): based on the user's online record in the week; Calculate and represent when life cycle, parameter value was configured to t, in the useless resources occupation rate of all inner filtration tables.

Step (5.2.2): if

less than explanation when choosing t as life cycle during parameter value; The useless resources occupation rate of filter table when choosing optimal is little; Therefore optimal is set as t again, simultaneously

is set as

again

Step (5.2.3): if

Equal

Then explanation is when choosing t and optimal as life cycle during parameter value, from useless resources occupation rate (R _W) this index, the two is of equal value, considers average number of duplicate acknowledgments (Avg this moment once more _DC) this filter table operating index.Therefore carry out following steps successively:

Step (5.2.3.1): based on the user's online record in the week; Calculate respectively

and wherein and

represented respectively when life cycle, parameter value was configured to t and optimal, one all in average number of duplicate acknowledgments.

Step (5.2.3.2): if

Less than

Then description selection t is as life cycle during parameter value, from average number of duplicate acknowledgments (Avg _DC) this index is better than and chooses optimal.Therefore optimal is set as t again.

Step (6): return the optimal that obtains initial optimum life cycle of configuration parameter value, and this value is configured in this SAVI switch through auto-configuration script as this SAVI switch.

2, adjust filter table parameter life cycle in real time

This part is achieved in the ageng of every SAVI switch itself:

Step (1): the keeper is provided with following parameter through configuration:

w _n: the weight that influences filtering meter item mean value;

C: the total capacity of filter table;

γ: adjustment threshold value (percentage of total capacity);

Max _Lifetime: the desirable maximum of configuration parameter life cycle;

Min _Lifetime: the desirable minimum value of configuration parameter life cycle;

p _Sampling: the sampling period (minute to be unit);

Optimal _Init: initial value is chosen initial optimum parameter value life cycle that the stage calculates;

Count _Max: the average list item number of this switch filter table surpasses the threshold value of C number of threshold value beta in the week.

Step (2): the following parameter of initialization:

Optimal _Cur: be illustrated in optimal value, it is set as optimal when parameter interior life cycle of the last week _Init.

Lifetime: parameter value life cycle that the SAVI switch should be chosen during current sampling point is set as optimal with it _Init

Count: the average filtration list item number of this switch in a week is set as 0 more than or equal to the number of times of threshold gamma C with it.

Step (3): in the switch actual moving process, the sampled point s for arriving, carry out following steps successively:

Step (3.1): if s is first sampled point in a certain week, but be not switch first sampled point in when operation, then carry out following steps successively:

Step (3.1.1): if count is less than count _Max, and optimal _CurLess than Max _Lifetime, then explanation is found through all actual motions on this switch, original envision low of the duty ratio of this switch filter table is therefore with optimal _CurAdjust upward a sampling period p _Sampling, be about to optimal _CurAgain be set as former optimal _CurAdd p _SamplingBut, optimal _CurCan not surpass Max _LifetimeAnd count is reinitialized into 0.

Step (3.1.2): if count is more than or equal to count _Max, and optimal _CurGreater than Min _LifetimeThen explanation finds that through all actual motions on this switch wanting of the original anticipation of the duty ratio of this switch filter table is high, therefore with optimal _CurAdjust a sampling period p downwards _Sampling, be about to optimal _CurAgain be set as former optimal _CurDeduct p _SamplingBut, optimal _CurCan not be lower than Min _LifetimeAnd count is reinitialized into 0.

Step (3.2): if s is switch first sampled point in when operation; Expression filter table parameter configuration life cycle becomes lifetime, the filter table item number when sampled point s then to make the average list item of filter table count that avg equals

.

Step (3.3): make the average list item of filter table count avg and equal following result: 1 deducts w _nAfter the avg that calculates when multiply by a sampled point, add w at last _nMultiply by

Step (3.4): if the average list item of filter table is counted avg greater than threshold gamma C; Explanation is at sampled point s; The load of filter table is heavier; Need rapid value minimizing, thereby let expired useless filtering meter item deleted as early as possible, thereby reduce the load of filter table configuration parameter lifetime life cycle.Therefore the value of lifetime is adjusted into following value: C and deducts behind the avg, again result and current lifetime value are multiplied each other at last again divided by C.And the value of this lifetime is set in parameter life cycle, if this result is less than Min _Lifetime, then the value with lifetime is adjusted into Min _LifetimeSimultaneously count is added 1.

Step (3.5): if the average list item of filter table is counted avg less than threshold gamma C, and lifetime is less than optimal _Cur, then lifetime is adjusted upward a sampling period, soon lifetime is set as former lifetime again and adds p _SamplingAnd the value of this lifetime is set in parameter life cycle.

Step (4): return step (3), get into the next sampling period.

SAVI switch network management system in Tsing-Hua University's campus network nearly the SAVI switch of 1000 actual motions monitor, periodically obtain the content of filter table in the switch.Because the students' dormitory of Tsing-Hua University adopts direct-connected mode to dispose the SAVI switch; Each port can detect the down incident that main frame breaks off network; Thereby delete useless filtering meter item fast; Therefore through the periodicity of SAVI switch filter table content is obtained, can collect user's internet behavior accurately.

Accurate internet information with the campus network user is that example is assessed the present invention below.

Based on the actual user's internet records that the collects simulation that makes an experiment; Draw to draw a conclusion from result of the test: although the list item number in the ideal case the filter table seldom; If but adopt cascade system to dispose the SAVI switch; And life cycle, parameter was configured to give tacit consent to filter table 4 hours, and so actual filter table item number is very many.If yet the self-adapting regulation method that uses the embodiment of the invention to provide, the load meeting is significantly reduced.

Obtain user's internet information 909 SAVI switches in campus network, these switches are distributed in about 30 sub-net.Utilize the approximate internet records accurately of 4 all users to simulate the cascade deployment way: to suppose to connect a virtual SAVI switch on current per three SAVI switches; These three SAVI switches think that then it is a general switch, and the filtering meter item of promptly original three switches will concentrate on this virtual SAVI switch.If the switch number in the subnet can not be divided exactly by 3, the then remaining corresponding switch of remainder will be fictionalized the SAVI switch of a cascade.

The filter table capacity of the SAVI switch of the SG2400 of StarNet model is 220, and a switch can direct-connected 24 main frames.Show that from user's internet information of collecting when using the IPv6 network, a main frame on average can produce the IPv6 address of three uses.If select 3 aggregate number during as experimental simulation so, in the ideal case, the number of users that connects under the SAVI switch that fictionalizes the chances are 216 (3 * 3 * 24).Therefore think and select 3 to be rational as aggregate number.The final SAVI switches that obtained 310 virtual cascade deployment altogether, and obtained their 4 corresponding all user's internet records.

At first ideal situation is defined.The SAVI switch adopts a kind of desirable mode to know accurately rolling off the production line the time of user, so switch can in time be deleted useless filtering meter item and farthest utilizes the filter table resource.Under the cascade deployment mode, message informs that to switch it leaves only if main frame initiatively sends notice, and more than the ideal situation of definition can not be implemented, but can use the adjustment target of this situation as other method.

Real user internet information based on 4 weeks is simulated the running of 310 virtual SAVI switches in the ideal case; The situation that also filter table parameter life cycle is statically configured as 15 minutes and 240 minutes is respectively simulated, and uses dynamic adjusting method to simulate the running of virtual switch at last.For dynamic adjusting method, the input parameter value is set is: w _n=0.5, C=220, β=0.8, Max _Lifetime=240 minutes, p _Sampling=5 minutes; At the second portion of method the input parameter value being set is: w _n=0.5, C=220, γ=0.5, Max _Lifetime=240 minutes, p _Sampling=5 minutes.

Selecting 240 minutes reasons as a kind of static configuration situation is that this value is the default value of the SG2400 of StarNet model switch filter table parameter life cycle; When other campus network adopts the cascade deployment way; If the keeper does not carry out any modification to this parameter, the actual motion effect of the switch of this type in these networks so.If selecting 15 minutes reasons as a kind of static configuration situation is to find through test; When adopting traditional static configuration method, have only the life cycle parameter value to be configured to be less than or equal to 15 minutes ability and to guarantee that any time of 310 SAVI switches in 4 weeks that fictionalizes put filter table and can not be filled.

Use R _W, Avg _DC, R _OAnd 4 indexs of filter table item number are assessed the operation conditions of virtual switch when adopting distinct methods.Calculate every interior R of virtual 4 weeks of SAVI switch _W, Avg _DCWith the value of filter table item number and draw respectively they cumulative probability distribution curve (cumulative distribution function, CDF).Fig. 3 is the CDF curve of filter table item number, and Fig. 4 is index Avg _DCThe CDF curve, Fig. 5 is index R _WThe CDF curve, the mean value of 310 virtual each indexs of SAVI switch is as shown in table 1.

Method	Life cycle parameter value	R O(%)	Avg DC	R W(%)	The average filtration table size
						Ideal situation	--	0	--	--	13.6
Static configuration	15 minutes	0	3.9	7.897	13.8
						Static configuration	240 minutes	1.106	0.085	73.61	61
Dynamically adjustment	--	0	0.439	55.89	28

Table 1

Experimental result shows:

Life cycle, parameters of choice can produce very big influence to the use of filtering meter item.Can see in the ideal case from Fig. 3 almost all being lower than 60, explain that under same virtual SAVI switch the user of networking seldom simultaneously at any one sampling period inner filtration list item number.But when parameter value is configured to give tacit consent to life cycle 240 minutes the time, surpass for 10% sampling period, the filter table item number of virtual SAVI switch surpasses 160.Can find out R when life cycle, parameter value was configured to 240 minutes through table 1 _O=1.106%, mean that nearly 15 minutes some terminal use can't use the IPv6 network in one day, so the default configuration value of equipment vendors is inappropriate.

The self-adapting regulation method that the present invention proposes can guarantee that the filter table resource is not wasted.Can also see that from Fig. 3 the filter table item number is lower than 120, can find out that through table 1 method of adjustment of the present invention can guarantee R simultaneously when using dynamic adjusting method of the present invention _O=0.

The self-adapting regulation method that the present invention proposes can guarantee the number of duplicate acknowledgments of filtering meter item few as much as possible.Although from the distribution of filter table item number when the method for using static configuration when life cycle, parameter configuration became 15 minutes, the CDF curve is near the curve under the ideal situation, from Fig. 4, Avg _DCThe CDF curve can find out, if with life cycle parameter configuration to become will cause in 15 minutes switch be to surpass 20% filtering meter item to surpass 6 times affirmation number of times, can find out Avg at this moment through table 1 simultaneously _DCNearly 4, mean that each filtering meter item of switch average out to carries out 5 times affirmation process, because the affirmation process not only will be sent message and main frame and carried out alternately.Switch also need be for each affirmation process be provided with a state machine simultaneously, and this will increase the running load of SAVI switch greatly.When using method of adjustment of the present invention, about 80% filtering meter item Avg _DC=0, promptly there is 80% filtering meter item need not carry out the duplicate acknowledgment process approximately.Can find out the average A vg of 310 virtual SAVI switches through table 1 _DCHave only 0.4; Mean that switch is on average as long as carry out 1.5 times affirmation process for each filtering meter item; Therefore the self-adapting regulation method of the embodiment of the invention is guaranteeing under the prerequisite that filter table is not filled, and has lowered the load that switch is confirmed process greatly.

Although when using static configuration method Avg when life cycle, parameter configuration became 240 minutes _DCVery little, but the filter table resource is but by serious waste.As can beappreciated from fig. 5, the useless resources occupation rate of its filter table of switch above 80% has surpassed 0.7, means that many filter table resources are wasted, and the poor efficiency utilization to filter table causes under the situation of this configuration, R can occurring just _O>0, promptly the situation of denial of service can appear in network.

In above analysis, compared various collocation method; Thereby draw to draw a conclusion: switch can not guarantee the availability of network, the low load and the use of filter table reasonable resources of switch simultaneously when using static configuration method; And dynamic adjusting method of the present invention can make switch operate under the good state, because the present invention has realized the good compromise of filter table item number and duplicate acknowledgment number.

In order further to understand the implementation of dynamic adjusting method; Choose two virtual SAVI switches and write down its filter table item number and configuration parameter value life cycle situation over time at Fig. 6 and Fig. 7 respectively; Wherein the load of a virtual SAVI switch is higher relatively, thereby as can beappreciated from fig. 6 the filter table item number often can surpass the adjustment threshold value and causes that the rapid adjustment of parameter life cycle deals with this emergency case; The load of another switch is then lighter relatively.Also the situation of two certain short periods of switch is amplified,, can more obviously find out parameter life cycle real-time change how like this along with the variation of filter table item number like Fig. 8 and shown in Figure 9.Life cycle incipient stage of adjustment first that parameter value is method to every initial optimal value that switch computes is come out.

Through analyzing, can draw 2 conclusions:

1) can reflect the actual conditions of every switch really in the initial optimal value of calculating of the self-adapting regulation method of the embodiment of the invention; If promptly pass through the data analysis in a week; Can a less relatively initial optimal value be set for parameter life cycle of the bigger switch of load; To guarantee that its useless filtering meter item can be by deletion as early as possible; And be that parameter life cycle of the lighter switch of load is provided with a relatively large initial optimal value, as much as possible little to guarantee its number of duplicate acknowledgments.

2) self-adapting regulation method of the embodiment of the invention is in the sudden growth that can catch the filter table item number when life cycle, parameter was adjusted in real time timely; And less parameter value life cycle is set, parameter value life cycle that increases after finishing when the burst of filtering the list item number can increase again gradually.

SAVI switch filter table parameter adaptive adjusting method life cycle according to the embodiment of the invention; Can be under the deployment way of SAVI switch in non-cascade; Variation according to filter table item number in the SAVI switch is dynamically adjusted filtering meter item parameter life cycle, is guaranteeing to reduce the number of duplicate acknowledgments of switch to filtering meter item as far as possible under the prerequisite that the switch filter table can not be filled at any time.The self-adapting regulation method of SAVI switch filter table parameter life cycle of the present invention can have been realized filter table item number decreased average 53%, and the number of duplicate acknowledgments to filtering meter item has reduced 92% simultaneously, has reached expection invention target.

Describe and to be understood that in the flow chart or in this any process otherwise described or method; Expression comprises module, fragment or the part of code of the executable instruction of the step that one or more is used to realize specific logical function or process; And the scope of preferred implementation of the present invention comprises other realization; Wherein can be not according to order shown or that discuss; Comprise according to related function and to carry out function by the mode of basic while or by opposite order, this should be understood by the embodiments of the invention person of ordinary skill in the field.

In flow chart the expression or in this logic of otherwise describing and/or step; For example; Can be considered to be used to realize the sequencing tabulation of the executable instruction of logic function; May be embodied in any computer-readable medium; Use for instruction execution system, device or equipment (like computer-based system, comprise that system or other of processor can be from the systems of instruction execution system, device or equipment instruction fetch and execution command), or combine these instruction execution systems, device or equipment and use.With regard to this specification, " computer-readable medium " can be anyly can comprise, storage, communication, propagation or transmission procedure are for instruction execution system, device or equipment or combine these instruction execution systems, device or equipment and the device that uses.The example more specifically of computer-readable medium (non-exhaustive list) comprises following: the electrical connection section (electronic installation) with one or more wirings; Portable computer diskette box (magnetic device); Random-access memory (ram), read-only memory (ROM) can be wiped and can edit read-only memory (EPROM or flash memory); Fiber device, and portable optic disk read-only memory (CDROM).In addition; Computer-readable medium even can be paper or other the suitable media that to print said program above that; Because can be for example through paper or other media are carried out optical scanner; Then edit, decipher or handle to obtain said program with other suitable methods in case of necessity with the electronics mode, then it is stored in the computer storage.

Should be appreciated that each several part of the present invention can use hardware, software, firmware or their combination to realize.In the above-described embodiment, a plurality of steps or method can realize with being stored in the memory and by software or firmware that suitable instruction execution system is carried out.For example; If realize with hardware; The same in another embodiment, each in the available following technology well known in the art or their combination realize: have the discrete logic that is used for data-signal is realized the logic gates of logic function, have the application-specific integrated circuit (ASIC) of suitable combinational logic gate circuit; Programmable gate array (PGA), field programmable gate array (FPGA) etc.

Those skilled in the art are appreciated that and realize that all or part of step that the foregoing description method is carried is to instruct relevant hardware to accomplish through program; Described program can be stored in a kind of computer-readable recording medium; This program comprises one of step or its combination of method embodiment when carrying out.

In addition, each functional unit in each embodiment of the present invention can be integrated in the processing module, also can be that the independent physics in each unit exists, and also can be integrated in the module two or more unit.Above-mentioned integrated module both can adopt the form of hardware to realize, also can adopt the form of software function module to realize.If said integrated module realizes with the form of software function module and during as independently production marketing or use, also can be stored in the computer read/write memory medium.

The above-mentioned storage medium of mentioning can be a read-only memory, disk or CD etc.

In the description of this specification, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means the concrete characteristic, structure, material or the characteristics that combine this embodiment or example to describe and is contained at least one embodiment of the present invention or the example.In this manual, the schematic statement to above-mentioned term not necessarily refers to identical embodiment or example.And concrete characteristic, structure, material or the characteristics of description can combine with suitable manner in any one or more embodiment or example.

Although illustrated and described embodiments of the invention; For those of ordinary skill in the art; Be appreciated that under the situation that does not break away from principle of the present invention and spirit and can carry out multiple variation, modification, replacement and modification that scope of the present invention is accompanying claims and be equal to and limit to these embodiment.

Claims (10)

1. SAVI switch filter table parameter adaptive adjusting method life cycle is characterized in that, comprises the steps:

Filter table parameter life cycle of SAVI switch is set, in said filter table parameter life cycle, collects the load condition of filter table of user's online record and the said SAVI switch of said SAVI switch in preset time work period;

According to the user's online record of said SAVI switch and the load condition of filter table, obtain a plurality of alternative filter table parameter values life cycle;

From said a plurality of alternative filter table parameter values life cycle, choose the inceptive filtering hypergene and deposit the period parameters optimal value, and filter table item number adjustment threshold value is set;

First sampled point in each said time work period that said SAVI switch moves based on the ruuning situation of a last work period temporal filtering table, calculates the filter table parameter life cycle optimal value of time this work period; And

Each sampled point in said time work period; Calculate the average list item number of said filter table, and the average list item number and the filter table item number adjustment threshold value of said filter table compared, and adjust said filter table parameter life cycle in real time according to comparative result; Wherein, At current sampling point, when the average list item number of said filter table is adjusted threshold value greater than said filter table item number, current filter table parameter value life cycle is turned down rapidly; Otherwise; If at current sampling point; When the average list item number of said filter table is adjusted threshold value less than said filter table item number; And filter table parameter life cycle of a last sampled point is during less than the optimal value of said filter table parameter life cycle, current filter table parameter value life cycle is transferred big, but this value does not allow the optimal value above said filter table parameter life cycle.

2. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 1; It is characterized in that; Said filter table parameter life cycle that the SAVI switch is set, filter table parameter life cycle that comprises the steps: to be provided with said SAVI switch is 5 minutes.

3. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 1 is characterized in that said preset time work period is 7 days.

4. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 1 is characterized in that, the said filter table item number adjustment threshold value that is provided with also further comprises:

The weight that influences filtering meter item mean value, the total capacity of filter table, maximum, the sampling period that life cycle, parameter allowed are set, and the first threshold number of times.

5. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 4 is characterized in that, initial optimum filter table parameter life cycle of said calculating comprises the steps:

For all filter table parameter values life cycle that allow and can be divided exactly by the sampling period, do following processing respectively:

Each sampled point in preset time work period; The average list item number of the filter table number of times that exceeds the quata is set to 0; Wherein, The average list item number of the said filter table number of times that exceeds the quata is used to add up said SAVI switch under the situation that is provided with of current filter table parameter value life cycle, and the average filtration list item number in preset time work period multiply by the number of times of filter table total capacity more than or equal to the adjustment threshold value;

Each sampled point in preset time work period calculates the average list item number of said filter table, when the average list item number of said filter table during greater than filter table item number adjustment threshold value, the average list item number of the filter table number of times that exceeds the quata is added 1;

If the situation that is provided with for current filter table parameter life cycle; For all sampled points in preset time work period; The average list item number of filter table exceeds the quata number of times less than preset said first threshold number of times, then lists current filter table parameter life cycle in initial optimum filter table parameter life cycle one of alternative;

According to the average number of duplicate acknowledgments of said preset work period temporal filtering table and the useless resources occupation rate of filter table, from the chosen candidate value of initial optimum filter table parameter life cycle, select the chosen candidate value of initial optimum filter table parameter life cycle of conduct.

6. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 5; It is characterized in that; First sampled point in each said time work period that said SAVI switch moves calculates said filter table parameter life cycle optimal value, also further comprises:

The average list item number of the said filter table number of times that exceeds the quata is reset to 0.

7. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 6 is characterized in that, the filter table parameter life cycle optimal value in said calculating work at present cycle comprises:

When said SAVI switch in last time work period; The average list item number of filter table exceeds the quata number of times less than preset said first threshold number of times; And during the maximum that the optimal value of said filter table parameter life cycle allows less than said filter table parameter life cycle in last time work period; The optimal value of last work period filter table parameter life cycle is added a time in sampling period, as the filter table parameter life cycle optimal value in work at present cycle.

8. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 6 is characterized in that, the filter table parameter life cycle optimal value in said calculating work at present cycle comprises:

When said SAVI switch in last time work period; The average list item number of filter table exceeds the quata number of times more than or equal to preset said first threshold number of times; And during the minimum value that the optimal value of said filter table parameter life cycle allows greater than said filter table parameter life cycle in last time work period; The optimal value of last work period filter table parameter life cycle is deducted a time in sampling period, as the filter table parameter life cycle optimal value in work at present cycle.

9. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 6; It is characterized in that; Said average list item number and filter table item number adjustment threshold value with said filter table compares, and adjusts said filter table parameter life cycle in real time according to comparative result, comprising:

At current sampling point, when the average list item number of said filter table during, filter table parameter life cycle of current sampling point is set, and the average list item number of the filter table number of times that exceeds the quata is added 1 according to following formula greater than filter table item number adjustment threshold value;

Wherein, lifetime _sBe filter table parameter life cycle of current sampling point, C is the total capacity of said filter table, avg _sBe the current average list item number of filter table, lifetime _S-1Filter table parameter life cycle for a last sampled point.

10. SAVI switch filter table parameter adaptive adjusting method life cycle as claimed in claim 6; It is characterized in that; The average list item number and the filter table item number adjustment threshold value of said filter table are compared, and adjust said filter table parameter life cycle in real time, comprising according to comparative result:

At current sampling point; When the average list item number of said filter table is adjusted threshold value less than the filter table item number; And filter table parameter life cycle of a last sampled point is during less than the optimal value of said filter table parameter life cycle; Filter table parameter life cycle of a last sampled point is added a time in sampling period, be set to filter table parameter life cycle of current sampling point.

Images