CN103744502B - Determine method and the device of the threshold frequency of embedded system - Google Patents
Determine method and the device of the threshold frequency of embedded system Download PDFInfo
- Publication number
- CN103744502B CN103744502B CN201310751302.8A CN201310751302A CN103744502B CN 103744502 B CN103744502 B CN 103744502B CN 201310751302 A CN201310751302 A CN 201310751302A CN 103744502 B CN103744502 B CN 103744502B
- Authority
- CN
- China
- Prior art keywords
- frequency
- embedded system
- management cycle
- arrival event
- moment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention provides a kind of method and the device of determining the threshold frequency of embedded system, method wherein comprises the first arrival event number (the arrival event number of arrival event upper limit curve) and the second arrival event number (frequency of operation in history management cycle and the arrival event number of historical perspective window) that the current management cycle obtains embedded system respectively; The arrival event number of next management cycle is obtained according to the first arrival event number and the second arrival event number; The frequency of operation of the embedded system of next management cycle is obtained according to the arrival event number of next management cycle obtained; According to the threshold frequency of the frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation determination embedded system of next management cycle.Utilize method and the device of the threshold frequency of the determination embedded system of the invention described above, when event obtains complete process, threshold frequency can be obtained fast, reduce the complexity solving threshold frequency.
Description
Technical field
The present invention relates to the dynamic electric voltage technical field regulating embedded system, more specifically, relate to a kind of method and the device of determining the threshold frequency of embedded system, change frequency of operation by threshold frequency, thus regulate dynamic electric voltage.
Background technology
In real-time embedded equipment, in order to economize energy, improve the serviceable life of battery simultaneously, generally need to manage power supply.Current method for managing power supply comprises dynamic power management (DPM, DynamicPowerManagement) and the large class of Dynamic voltage scaling (DVS, DynamicVoltageScaling) two.
In dynamic power management, the duty of processor be divided into activity, dormancy and stop three kinds of states, according to the arrival state of required process event, in time processor being switched to activity or dormant state, ensure that event is processed in time on the one hand, saved the energy on the other hand.And the power management thinking of Dynamic voltage scaling, be then the load according to arrival event, suitably adjust the service voltage of processor, thus change its frequency of operation, the event that both ensured is processed in time, again saves the energy.
In Dynamic voltage scaling, in each voltage decision-making moment, according to the arrival event upper limit curve of event and the state of current processor, determine processor running frequency and the relevant voltage of subsequent time.In all kinds of Dynamic voltage scaling algorithm, OPT algorithm is known the best way that significantly can reduce energy consumption.Being described in detail as follows this algorithm:
At moment t, for an also untreated event e completed
jif, c
jt () is that it is at moment t and processor speed s
maxresidue in the situation the worst processing time.α
jthe time of arrival of this event, d
jit is the stop time point of this event.Then OPT algorithm can describe with following formula:
Wherein,
for the lowest operating frequency of processor, this value is according to processor characteristic, in order to meet energy saving requirement and some processor characteristics and the fixed value drawn.
OPT algorithm meets the low-limit frequency requirement processing all events with the way selection of greediness, thus reaches the object of economize energy; If the processing speed constant interval of processor is (0, + ∞), then OPT algorithm is the Dynamic voltage scaling algorithm of feasible optimum, but actual conditions as shown in Figure 1, the frequency of operation calculated with OPT algorithm likely can be greater than the maximal value that processor can reach, thus makes OPT algorithm infeasible, when the actual frequency of processor is lower than the value calculated, some events then can be caused can not to be finished before its stop time point, thus managing electrical power failure.
On the basis of OPT algorithm, by arranging a threshold frequency s
*, when the frequency s calculating gained is greater than this threshold frequency s
*time, s is directly adjusted to maximal value, thus ensures to there will not be the frequency s calculated to be greater than the situation of maximum frequency.As shown in Figure 2, operational effect as shown in Figure 3 for its choice algorithm.
Now, how threshold frequency s is selected
*become crucial, if s
*select too much, will still cause occurring that s that subsequent calculations goes out is greater than the situation of maximum frequency, if s
*select too small, processor long-play will be made at maximum frequency, lose the meaning of power management; At present to s
*selection, generally adopt trial and error, namely first select a value, then derive on this basis and whether can meet the situation that s that subsequent calculations goes out is not more than maximum frequency, if satisfied, more suitably increase s
*value, otherwise, then reduce s
*value, use the method generally can approx near desirable s
*value, but the method needs repeatedly to attempt, and solving speed is slower.
Therefore, a kind of brand-new calculating threshold frequency s is needed
*scheme.
Summary of the invention
In view of the above problems, the object of this invention is to provide a kind of method and the device of determining the threshold frequency of embedded system, need repeatedly to attempt to solve calculating threshold frequency, the problem that solving speed is slower.
The invention provides a kind of method determining the threshold frequency of embedded system, embedded system has management cycle T, and centered by the nT moment, the time period in (n-1) T moment to nT moment forms the current management cycle; The nT moment forms next management cycle to the time period in (n+1) T moment; N the management cycle between initial time to nT moment forms the history management cycle, n > 0; (n-m) m the management cycle between the T moment to nT moment forms historical perspective window, 0≤m < n;
The method comprises:
The first arrival event number and the second arrival event number of embedded system is obtained respectively in the current management cycle; Wherein, first arrival event number is the arrival event number according to obtaining the acquisition of arrival event upper limit curve, and the second arrival event number is obtain the frequency of operation in history management cycle and the arrival event number of historical perspective window according to the historical record of embedded system;
The arrival event number of next management cycle of current management cycle is obtained according to the first arrival event number and the second arrival event number;
The frequency of operation of the embedded system of next management cycle is obtained according to the arrival event number of next obtained management cycle;
According to the threshold frequency of the frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation determination embedded system of next management cycle.
The present invention also provides a kind of device determining the threshold frequency of embedded system, and embedded system has management cycle T, and the time period in (n-1) T moment to nT moment forms the current management cycle; The nT moment forms next management cycle to the time period in (n+1) T moment; N the management cycle between initial time to nT moment forms the history management cycle, n > 0; (n-m) m the management cycle between the T moment to nT moment forms historical perspective window, 0≤m < n;
This device comprises:
Arrival event number acquiring unit, for obtaining the first arrival event number and the second arrival event number of embedded system respectively in the current management cycle; Wherein, first arrival event number is the arrival event number obtaining arrival event upper limit curve according to the feature of embedded system, and the second arrival event number is obtain the frequency of operation in history management cycle and the arrival event number of historical perspective window according to the historical record of embedded system;
Next management cycle arrival event number acquiring unit, for obtaining the arrival event number of next management cycle of current management cycle according to the first arrival event number and the second arrival event number;
Frequency of operation acquiring unit, for obtaining the frequency of operation of the embedded system of next management cycle according to the arrival event number of next obtained management cycle;
Threshold frequency determining unit, for the threshold frequency according to the frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation determination embedded system of next management cycle.
Utilize above-mentioned according to method and the device of determining the threshold frequency of embedded system of the present invention, when event obtains complete process, threshold frequency can be obtained fast, reduce the complexity solving threshold frequency.
In order to realize above-mentioned and relevant object, will describe in detail and the feature particularly pointed out in the claims after one or more aspect of the present invention comprises.Explanation below and accompanying drawing describe some illustrative aspects of the present invention in detail.But what these aspects indicated is only some modes that can use in the various modes of principle of the present invention.In addition, the present invention is intended to comprise all these aspects and their equivalent.
Accompanying drawing explanation
By reference to the content below in conjunction with the description of the drawings and claims, and understand more comprehensively along with to of the present invention, other object of the present invention and result will be understood and easy to understand more.In the accompanying drawings:
Fig. 1 is the service chart that the frequency of operation of embedded system is greater than maximum frequency;
Fig. 2 is the process flow diagram of threshold frequency selection algorithm;
Fig. 3 is the service chart of threshold frequency;
Fig. 4 is the process flow diagram of the method for the threshold frequency of determination embedded system according to the embodiment of the present invention;
Fig. 5 ~ Fig. 8 is the process flow diagram specifically determining the threshold frequency of embedded system according to the embodiment of the present invention;
Fig. 9 is the building-block of logic of the device of the threshold frequency of determination embedded system according to the embodiment of the present invention.
Label identical in all of the figs indicates similar or corresponding feature or function.
Embodiment
In the following description, for purposes of illustration, in order to provide the complete understanding to one or more embodiment, many details have been set forth.But, clearly, also these embodiments can be realized when there is no these details.In other example, one or more embodiment for convenience of description, known structure and equipment illustrate in block form an.
First some relations and concept are made an explanation before addressing detailed embodiments of the invention:
The current management cycle: be made up of to the time period in nT moment (n-1) T moment.
Next management cycle; Be made up of the time period of nT moment to (n+1) T moment.
The history management cycle: be made up of n the management cycle initial time to nT moment.
Historical perspective window: be made up of m the management cycle (n-m) T moment to nT moment; Wherein, 0≤m < n;
Have foregoing to find out, historical perspective window is within the scope in history management cycle, that is, the management cycle number of historical perspective window is less than the number in history management cycle.
Below with reference to accompanying drawing, specific embodiments of the invention are described in detail.
Fig. 4 is the flow process of the method for the threshold frequency of determination embedded system according to the embodiment of the present invention.
As shown in Figure 4, the method for the threshold frequency of embodiment of the present invention determination embedded system comprises:
S401: the first arrival event number and the second arrival event number that obtain embedded system in the current management cycle respectively; Wherein, first arrival event number is the arrival event number obtained according to arrival event upper limit curve, and the second arrival event number is the arrival event number obtained according to the historical record of embedded system in the frequency of operation in history management cycle and historical perspective window.
First, be described and modeling event, select series of parameters to describe an event, these parameters are:
The worst execution time (worst-caseexecutiontime) of C---event;
The time of arrival (arrivaltime) of α---event;
The stop time point (deadline) of d---event, has crossed this stop time point, if event is not processed complete, then abandons this event.
In actual environment, the arrival of event may be erratic, random, the arrival of predicted events exactly be very difficult; Therefore, by definition arrival event upper limit curve, the arrival rule of event can be described approx.Sequence of events can by arrival event upper limit curve
describe, when
R (t) is the function of the arrival event number in t from the moment 0 to the moment;
be from moment t, in delta time section, event arrives the upper limit of number, therefore arrival event upper limit curve is called, this curve can pass through system features (descriptor or the mode by application typical case input) and obtain, such as, if an event set is periodic event collection, system features has just determined the arrival event upper limit curve of this periodic event collection, just can obtain the arrival event number of the arrival event upper limit curve of this periodic event collection.
And the frequency of operation obtaining the management cycle in history week is exactly the frequency of operation obtaining n the management cycle in the history management cycle respectively; The arrival event number obtaining historical perspective window is the arrival event number of m the management cycle obtained respectively in historical perspective window, respectively calculate (n-m) T moment to the nT moment, (n-m+1) T moment to nT moment ..., (n-1) T moment to the nT moment arrival event number, get the arrival event number of minimum value as historical perspective window of 3 arrival event numbers.
The present embodiment is preferred embodiment: m=3, namely 3 management cycles between (n-3) T moment to nT moment are selected to form historical perspective window, calculate (n-3) T moment to nT moment, (n-2) T moment to nT moment and the arrival event number in (n-1) T moment to nT moment respectively, get the arrival event number of minimum value as historical perspective window of 3 arrival event numbers.
S402: the arrival event number obtaining next management cycle of current management cycle according to the first arrival event number and the second arrival event number.
Wherein, the arrival event number that the arrival event number of next management cycle equals arrival event upper limit curve deducts the result of the arrival event number of historical perspective window.
S403: the frequency of operation obtaining the embedded system of next management cycle according to the arrival event number of next obtained management cycle.
Wherein, arrival event number and frequency of operation are corresponding, and arrival event number is more, and frequency of operation is higher; Arrival event is fewer, and frequency of operation is lower.
S404: according to the threshold frequency of the frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation determination embedded system of next management cycle.
Determine that the detailed process of the threshold frequency of embedded system is: judge whether the frequency of operation of next management cycle exceeds the maximum frequency of embedded system, if do not exceeded, then using the threshold frequency of the maximum frequency of embedded system as embedded system; If exceeded, then the arrival event corresponding to the part exceeding the maximum frequency of embedded system is advanced to history management period treatment, and using the threshold frequency of frequency of operation as embedded system processing the management cycle of all arrival events at the maximum frequency not exceeding described embedded system in the history management cycle.
After determining the threshold frequency of embedded system, judge whether the frequency of operation of next management cycle is more than or equal to the threshold frequency of embedded system, if be more than or equal to, then embedded system is run with its maximum frequency; If be less than, embedded system is run with the frequency of operation of next management cycle.
Above-mentioned steps describes the method for the threshold frequency determining embedded system, determine that this threshold frequency is determined by the mode of line, that is, by obtaining the service chart of embedded system, adopt the mode of line on service chart, mark the threshold frequency of embedded system.
By from three aspects, above-mentioned method is described in detail below.
One, event arrives prediction
As mentioned above, arrival event upper limit curve
for event in delta time section arrives the upper limit of number, management cycle is T, for every T time, needs frequency of operation and the operating voltage of the embedded system calculating next management cycle, therefore, the event arrived between (n-1) T moment to nT moment, all by considering its impact on system in the nT moment, therefore can think that event all arrives in the nT moment simultaneously, arrive in advance and will store in the buffer with pending, by that analogy, for the nT moment, according to arrival event upper limit curve
predict that its event in next management cycle (time period from the nT moment to (n+1) the moment) arrives the upper limit of number; If
then within the next management cycle, the upper limit of event arrival number is 10, i.e. n
nT, max(nT)=10.
On the other hand, operating voltage changes along with frequency of operation, frequency of operation becomes large then operating voltage and becomes large, frequency of operation diminishes, and operating voltage diminishes, by regulating the frequency of operation of each management cycle in embedded system thus regulating the operating voltage of each management cycle in embedded system, reach the object of power management.
Two, historical perspective window is introduced
If only with the prediction in the nT moment, the upper limit that the nT moment drawn arrives number to the event in (n+1) T time period in moment is too rough, to predict that the nT moment arrives the upper limit of number to the event in (n+1) T time period in moment further more accurately, suitably should expand prediction window, consider nT moment existing event number, and the event in each cycle before the nT moment arrives number situation, namely introduces historical perspective window; To observe (n-1) T moment, setting
(arrival event upper limit curve), therefore in time period between (n-1) T moment to (n+1) T moment, event arrives the upper limit of number is 15, the leave of absence number being the arrival of 10(historical perspective window due to the event number arrived in the time period between (n-1) T moment to nT moment is 10), therefore the upper limit that the event in time period between the nT moment to (n+1) T moment arrives number is 5, i.e. n
nT, max[(n-1) T]=5, therefore, the upper limit that event can be arrived number converges to a less value.
Same reason, when the historical perspective window selected continues to expand, can draw n respectively
nT, max[(n-2) T], n
nT, max[(n-3) T] ..., n
nT, max(0) value, then the upper limit that the event in the time period between the nT moment to (n+1) T moment arrives number is:
N
nT, max=min
0≤i<n{ (n
nT, max(iT)) } (formula 2)
Wherein, n
nT, max(0) represent that maximal value got by historical perspective window, be from the 0T moment to all management cycles the nT moment as historical perspective window, particularly, the historical perspective window size selected is 3T, namely calculate (n-3) T moment to nT moment, (n-2) T moment to nT moment and (n-1) T moment event number to the arrival in nT moment respectively, and get its minimum value historical perspective window the most.
Three, overclocking thought is introduced
Overclocking is a large feature of modern processors, namely processor can be operated on its maximum frequency, " overload " carry out work, such as, processor can only process 10 events within each cycle, and predict that arrival event has 15, then in order to process all events, processor just needs overclocking work.In embodiments of the present invention, the process event of embedded system is completed by the processor in embedded system, and the processor of the setting embodiment of the present invention can be operated in overclocking pattern, and the frequency of operation of processor belongs to interval (0 ,+∞).
Threshold frequency s in the embodiment of the present invention
*represent, solve s
*process be exactly that the arrival event corresponding to the part of overclocking work is processed, namely by arrival event corresponding for the part exceeding maximum frequency, the cycle before moving to it completes.
The threshold frequency of the embedded system of the embodiment of the present invention is determined by the mode of line, first the service chart of processor is obtained, horizontal ordinate in this service chart is each management cycle, ordinate is the frequency of processor, and each rectangle in service chart is the frequency of operation of each management cycle process event.
The event processed in each management cycle is classified, to be divided in the last management cycle in this moment of next-door neighbour newly arrived event (being called category-A event) and to be arrived before next-door neighbour's last management cycle in this moment and the event (being called category-B event) also do not processed to this moment, for the nT moment, the new events that namely category-A event arrives between (n-1) T moment and nT moment, and category-B event refers to and to arrive and the event still do not processed to the nT moment before (n-1) T moment, therefore, in each cycle T, the work of processor can be divided into two classes, namely category-A event and treatments B class event is processed, the order of setting process device process event first processes category-A event, and then treatments B class event, for the management cycle of processor overclocking work, the reason of its overclocking has two, an overclocking being category-A event and too much causing, another is the overclocking that category-B event too much causes, for the overclocking that category-A event too much causes, mean current period inner treater all for the treatment of newly arrived event, but still can not process newly arrived event completely when not overclocking, in this case the too many non-scheduling of event is described, our situation that consideration event can be scheduled, therefore, only consider that category-B event too much causes the situation of overclocking, because category-B event just arrives before being the last management cycle, therefore can be moved within the last management cycle and process, and the process of event is moved forward, still can meet the requirement of its stop time point, the like, if the last management cycle can process all events when not overclocking, then event reach process terminates, otherwise, still the event in the last management cycle can be moved forward, until event has processed completely, or obtains the situation of non-scheduling.
Fig. 5 ~ Fig. 8 shows the flow process specifically determining the threshold frequency of embedded system according to the embodiment of the present invention.
As shown in Figure 5, because only there are two work period overclockings in its event handler procedure, therefore process each overclocking part from back to front, the category-B event of overclocking part is advanced to upper one-period process, after reach as shown in Figure 6, now, still need to distinguish the event in the last management cycle, save differentiation process here, if the category-A event after distinguishing and category-B event ratio still remain unchanged, then the part of overclocking is continued reach, after reach as shown in Figure 7; By that analogy, as shown in Figure 8, final 7th management cycle reciprocal can process all events, s now
*move to the initial launch frequency of the 7th management cycle reciprocal.
Generally speaking, in overclocking processing procedure, as shown in Fig. 8, inverse the 6th cycle (before not mobile overclocking) overclocking part is put in the previous cycle and is gone, and that is, the previous cycle calculated a frequency s originally, should run by s in theory, but in order to process the additional event of rear one-period, needing the frequency by being greater than s to run, therefore s being set as s
*as long as the frequency calculated is more than or equal to s
*, just press maximum frequency and run, ensure that all events are processed in time with this.S
*the i.e. s value that originally calculated of overclocking previous cycle.
For the situation of repeatedly overclocking, all such process is carried out to each overclocking part, draws corresponding s respectively
*value, final s
*value is s
*in minimum value.
If processor does not have the situation of overclocking process, then threshold frequency s
*it is exactly maximum frequency.
Can when event obtains complete process by the method for the threshold frequency of the determination embedded system of the invention described above embodiment, accurately obtain threshold frequency, reduce the complexity solving threshold frequency, the frequency of operation of embedded system is suitably regulated by threshold frequency, thus suitably adjust the service voltage of processor, while the event of guarantee obtains process in time, again save the energy.
With said method relativity, the present invention also provides a kind of device determining the threshold frequency of embedded system, Fig. 9 shows the device of the threshold frequency of the determination embedded system according to the embodiment of the present invention, as shown in Figure 9, provided by the inventionly determine that the device 900 of the threshold frequency of embedded system comprises:
Arrival event number acquiring unit 910, for obtaining the first arrival event number and the second arrival event number of embedded system respectively in the current management cycle; Wherein, first arrival event number is the arrival event number obtaining arrival event upper limit curve according to the feature of embedded system, and the second arrival event number is obtain the frequency of operation in history management cycle and the arrival event number of historical perspective window according to the historical record of embedded system;
Next management cycle arrival event number acquiring unit 920, for obtaining the arrival event number of next management cycle of current management cycle according to the first arrival event number and the second arrival event number;
Frequency of operation acquiring unit 930, for obtaining the frequency of operation of the embedded system of next management cycle according to the arrival event number of next obtained management cycle;
Threshold frequency determining unit 940, for the threshold frequency according to the frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation determination embedded system of next management cycle.
Wherein, described threshold frequency determining unit 940 comprise arrival event reach module (scheming not shown), for the maximum frequency by exceeding embedded system part corresponding to arrival event be advanced to history management period treatment; Maximum frequency judge module (scheming not shown), for judging whether the frequency of operation of next management cycle exceeds the maximum frequency of embedded system, if do not exceeded, then threshold frequency determining unit is using the threshold frequency of the maximum frequency of embedded system as embedded system; If exceeded, then the arrival event corresponding to the part exceeding the maximum frequency of embedded system is advanced to history management period treatment by arrival event reach module, further, threshold frequency determining unit is using the threshold frequency of frequency of operation as embedded system processing the management cycle of all arrival events at the maximum frequency not exceeding described embedded system in the history management cycle.
This device also comprises described device and also comprises maximum frequency running unit (scheming not shown), for when the frequency of operation of next management cycle is more than or equal to the threshold frequency of embedded system, runs embedded system with its maximum frequency; Frequency of operation running unit (scheming not shown), for when the frequency of operation of next management cycle is less than the threshold frequency of embedded system, runs embedded system with the frequency of operation of next management cycle.
Method and the device of the threshold frequency of the determination embedded system proposed according to the present invention is described in an illustrative manner above with reference to accompanying drawing.But, it will be appreciated by those skilled in the art that method and the device of the threshold frequency of the determination embedded system that the invention described above is proposed, various improvement can also be made on the basis not departing from content of the present invention.Therefore, protection scope of the present invention should be determined by the content of appending claims.
Claims (7)
1. determine a method for the threshold frequency of embedded system, described embedded system has management cycle T, and centered by the nT moment, the time period in (n-1) T moment to nT moment forms the current management cycle; The nT moment forms next management cycle to the time period in (n+1) T moment; N the management cycle between initial time to nT moment forms the history management cycle, n > 0; (n-m) m the management cycle between the T moment to nT moment forms historical perspective window, 1≤m < n;
Described method comprises:
The first arrival event number and the second arrival event number of described embedded system is obtained respectively in the current management cycle; Wherein, described first arrival event number is the arrival event number obtained according to arrival event upper limit curve, and described second arrival event number is the arrival event number obtained according to the historical record of described embedded system in the frequency of operation in described history management cycle and described historical perspective window;
The arrival event number of next management cycle of described current management cycle is obtained according to described first arrival event number and described second arrival event number;
The frequency of operation of the described embedded system of next management cycle described is obtained according to the arrival event number of next management cycle described in obtained;
The threshold frequency of described embedded system is determined according to the described frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation of next management cycle described.
2. determine the method for the threshold frequency of embedded system as claimed in claim 1, wherein, determining in the process of the threshold frequency of described embedded system according to the described frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation of next management cycle described
Judge whether the frequency of operation of next management cycle described exceeds the maximum frequency of described embedded system; Wherein,
If the frequency of operation of next management cycle described does not exceed the maximum frequency of described embedded system, then using the threshold frequency of the maximum frequency of described embedded system as described embedded system;
If the frequency of operation of next management cycle described exceeds the maximum frequency of described embedded system, then the arrival event corresponding to the part exceeding the maximum frequency of described embedded system is advanced to history management period treatment, and using the threshold frequency of frequency of operation as described embedded system processing the management cycle of all arrival events at the maximum frequency not exceeding described embedded system in the history management cycle.
3. determine the method for the threshold frequency of embedded system as claimed in claim 1, wherein,
After the threshold frequency determining described embedded system, judge whether the frequency of operation of next management cycle described is more than or equal to the threshold frequency of described embedded system, if be more than or equal to, then described embedded system is run with its maximum frequency; If be less than, described embedded system is run with the frequency of operation of next management cycle described.
4. determine the method for the threshold frequency of embedded system as claimed in claim 1, wherein,
3 management cycles between (n-3) T moment to nT moment are selected to form historical perspective window, calculate (n-3) T moment to nT moment, (n-2) T moment to nT moment and the arrival event number in (n-1) T moment to nT moment respectively, get the arrival event number of minimum value as historical perspective window of 3 arrival event numbers.
5. determine a device for the threshold frequency of embedded system, described embedded system has management cycle T, and centered by the nT moment, the time period in (n-1) T moment to nT moment forms the current management cycle; The nT moment forms next management cycle to the time period in (n+1) T moment; N the management cycle between initial time to nT moment forms the history management cycle, n > 0; (n-m) m the management cycle between the T moment to nT moment forms historical perspective window, 1≤m < n;
Described device comprises:
Arrival event number acquiring unit, for obtaining the first arrival event number and the second arrival event number of described embedded system respectively in the current management cycle; Wherein, described first arrival event number is the arrival event number obtained according to arrival event upper limit curve, and described second arrival event number is obtain the frequency of operation in described history management cycle and the arrival event number of described historical perspective window according to the historical record of described embedded system;
Next management cycle arrival event number acquiring unit, for obtaining the arrival event number of next management cycle of current management cycle according to described first arrival event number and described second arrival event number;
Frequency of operation acquiring unit, for obtaining the frequency of operation of the embedded system of next management cycle according to the arrival event number of next obtained management cycle;
Threshold frequency determining unit, for determining the threshold frequency of described embedded system according to the described frequency of operation in history management cycle, the maximum frequency of embedded system and the frequency of operation of next management cycle described.
6. determine the device of the threshold frequency of embedded system as claimed in claim 5, wherein,
Described threshold frequency determining unit comprises:
Arrival event reach module, for the maximum frequency by exceeding described embedded system part corresponding to arrival event be advanced to history management period treatment;
Maximum frequency judge module, for judging whether the frequency of operation of next management cycle described exceeds the maximum frequency of described embedded system, if do not exceeded, then described threshold frequency determining unit is using the threshold frequency of the maximum frequency of described embedded system as described embedded system; If exceeded, then the arrival event corresponding to the part exceeding the maximum frequency of described embedded system is advanced to history management period treatment by arrival event reach module, further, described threshold frequency determining unit is using the threshold frequency of frequency of operation as described embedded system processing the management cycle of all arrival events at the maximum frequency not exceeding described embedded system in the history management cycle.
7. determine the device of the threshold frequency of embedded system as claimed in claim 5, wherein,
Described device also comprises:
Maximum frequency runs module, for when the frequency of operation of next management cycle described is more than or equal to the threshold frequency of described embedded system, runs described embedded system with its maximum frequency;
Frequency of operation runs module, for when the frequency of operation of next management cycle described is less than the threshold frequency of described embedded system, runs described embedded system with the frequency of operation of next management cycle described.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310751302.8A CN103744502B (en) | 2013-12-31 | 2013-12-31 | Determine method and the device of the threshold frequency of embedded system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310751302.8A CN103744502B (en) | 2013-12-31 | 2013-12-31 | Determine method and the device of the threshold frequency of embedded system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103744502A CN103744502A (en) | 2014-04-23 |
| CN103744502B true CN103744502B (en) | 2016-01-27 |
Family
ID=50501527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310751302.8A Active CN103744502B (en) | 2013-12-31 | 2013-12-31 | Determine method and the device of the threshold frequency of embedded system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103744502B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1784646A (en) * | 2003-05-07 | 2006-06-07 | 苹果电脑公司 | Method and apparatus for dynamic power management in a processor system |
| CN101226421A (en) * | 2008-01-16 | 2008-07-23 | 浙江大学 | MSR method for real time embedded system EDF low-power consumption scheduling |
| CN102207769A (en) * | 2011-05-24 | 2011-10-05 | 东北大学 | Static voltage scheduling-based energy optimization method of DVS (Dynamic Voltage Scaling) system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2446830B (en) * | 2007-02-22 | 2009-08-26 | Toshiba Res Europ Ltd | Controller for processing apparatus |
-
2013
- 2013-12-31 CN CN201310751302.8A patent/CN103744502B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1784646A (en) * | 2003-05-07 | 2006-06-07 | 苹果电脑公司 | Method and apparatus for dynamic power management in a processor system |
| CN101226421A (en) * | 2008-01-16 | 2008-07-23 | 浙江大学 | MSR method for real time embedded system EDF low-power consumption scheduling |
| CN102207769A (en) * | 2011-05-24 | 2011-10-05 | 东北大学 | Static voltage scheduling-based energy optimization method of DVS (Dynamic Voltage Scaling) system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103744502A (en) | 2014-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9459918B2 (en) | Scheduling threads | |
| CN108984292B (en) | Fixed priority periodic task energy consumption optimization method for hybrid key system | |
| US20140344599A1 (en) | Method and System for Power Management | |
| CN108521837B (en) | Micro-energy acquisition-based power supply management method and device and micro-energy power supply | |
| JP2013536532A5 (en) | ||
| JP2012506597A5 (en) | ||
| KR20160147930A (en) | Load scheduling in multi-battery devices | |
| EP3855286B1 (en) | Dormancy controlling method for on board computing platform, device and readable storage medium | |
| RU2013119123A (en) | APPLICATION LIFE CYCLE MANAGEMENT | |
| CN101103336A (en) | Data processing system and method of task scheduling | |
| CN105608049A (en) | Method and device for controlling CPU of intelligent terminal | |
| CN107878366A (en) | A kind of method and device for being used to control new energy vehicle controller working condition | |
| CN103365710A (en) | Real-time task scheduling device and method and computer system | |
| CN105677000A (en) | Dynamic voltage and frequency scaling system and method | |
| CN106445070A (en) | Energy consumption optimization scheduling method for hard real-time system resource-limited sporadic tasks | |
| CN105320570A (en) | Resource management method and system | |
| CN106681475A (en) | Method for reducing power consumption of mobile terminal and mobile terminal | |
| CN106124835B (en) | A kind of method that power distribution voltage compares with predetermined voltage | |
| CN103744502B (en) | Determine method and the device of the threshold frequency of embedded system | |
| Duque et al. | Energy consumption control of one machine manufacturing system with stochastic arrivals based on fuzzy logic | |
| CN103823544B (en) | A kind of dynamic power supply management method | |
| CN105630126A (en) | Low-power scheduling method for mixed task based on constant bandwidth server | |
| CN104915559B (en) | A kind of electrical equipment determines method and system | |
| CN105022469A (en) | Portable electronic device and core swapping method thereof | |
| CN103645797B (en) | A kind of low power consumption frequency modulation based on timeslice utilization factor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |