CN110769490A

CN110769490A - Intelligent grouping self-adaptive energy-saving method suitable for multiple terminal systems

Info

Publication number: CN110769490A
Application number: CN201910813472.1A
Authority: CN
Inventors: 樊祥宁; 曹国锦; 尹喜阳; 胡庆生; 岳顺民; 郭延凯
Original assignee: Southeast University; State Grid Tianjin Electric Power Co Ltd
Current assignee: Southeast University; State Grid Tianjin Electric Power Co Ltd
Priority date: 2019-08-30
Filing date: 2019-08-30
Publication date: 2020-02-07
Anticipated expiration: 2039-08-30
Also published as: CN110769490B

Abstract

The invention discloses an intelligent grouping self-adaptive energy-saving method suitable for various terminal systems, which is characterized in that for a system comprising various intelligent terminals, the terminals are intelligently grouped, and corresponding default values are configured for each group; judging whether the configuration parameters need to be optimized or not; if optimization is needed, self-adaptively adjusting and optimizing parameter configuration through DRX + eDRX, and determining DRX and/or eDRX working cycle; if the optimization is not needed, continuing to execute the current configuration, and entering energy-saving data statistics; judging whether the parameters need to be reconfigured according to the energy-saving effect; if the parameters need to be reconfigured, the working period is adaptively adjusted; if the parameters do not need to be reconfigured, judging whether a new terminal exists; if a new terminal exists, intelligently grouping the terminals and configuring default values; and if no new terminal exists, continuing to execute the current operation, and performing energy-saving statistics after a period of time. According to the method and the device, the multi-gear DRX period and the multi-gear eDRX period are freely switched, so that the energy consumption of the system is reduced.

Description

Intelligent grouping self-adaptive energy-saving method suitable for multiple terminal systems

Technical Field

The invention relates to the technical field of application of NB-IoT in a smart power grid, in particular to an intelligent grouping self-adaptive energy-saving method suitable for various terminal systems.

Background

In the idle state, a maximum value of a Discontinuous Reception (DRX) cycle widely used in LTE is 2.56s, and frequent wake-up consumes energy of the terminal. To further reduce power consumption and extend sleep time without important paging information, R13 introduces an extended discontinuous reception (eDRX) pattern in NB-IoT. The terminal sleep time is increased on the basis of DRX discontinuous paging so as to reduce total energy consumption.

However, the simple DRX cycle adjustment can respond to paging in time due to more paging times, but may waste resources; pure eDRX cycle adjustment may have sufficient rest time due to the long sleep times at the end of each cycle, but may not allow timely response to paging.

A plurality of intelligent terminals such as intelligent electric meters, water meters, gas meters and the like are arranged around the intelligent electric meter. Generally, an intelligent terminal has a predetermined reporting period. Based on the fact that the number of the intelligent terminals is large, and the terminals in different types of places have different reporting requirements. If all terminals respond at the same time, it is clear that the channel becomes very congested at that moment and that no timely response to paging is possible.

In the face of huge number of intelligent terminals, if all the processes are completely operated by manpower, a large amount of manpower and material resources are wasted. Most of the existing intelligent systems are fully automatic, and the addition of the terminal, the report of the terminal and the like are automatically carried out by the system. At this time, if there is no operation for timely judgment, the system may be confused.

Disclosure of Invention

The technical problem to be solved by the invention is as follows:

the invention provides an intelligent grouping self-adaptive energy-saving method suitable for various terminal systems, which aims to overcome the defects caused by the independent use of the existing DRX and eDRX, solve the problems caused by the simultaneous response of multiple terminals to channels, responses and the like and match with the automatic operation characteristics of a system.

The invention adopts the following technical scheme for solving the technical problems:

the invention provides an intelligent grouping self-adaptive energy-saving method suitable for various terminal systems, which comprises the following steps:

the method comprises the following steps that firstly, intelligent grouping is carried out on all terminals, and default values are configured for each group of terminals;

step two, judging whether the optimal configuration is needed or not according to the grouping result; if the optimal configuration is needed, entering a third step; otherwise, entering the step four;

step three, optimizing configuration: the working cycle is adaptively adjusted through DRX + eDRX, and the working cycle of DRX and/or eDRX is finally determined;

step four, energy-saving statistics is carried out, and if the energy-saving effect does not reach the preset standard, triple new optimization configuration is carried out;

step five, judging whether a new terminal is added; if a new terminal is added, returning to execute the step one; and if no new terminal is added, continuing to execute the current operation, and performing energy-saving statistics after a period of time.

The foregoing method for intelligent grouping and adaptive energy saving suitable for multiple terminal systems, further, the step of intelligently grouping terminals includes:

step 1.1, preprocessing the reporting time of each terminal: enabling the terminal to work for a period of time, and collecting the reporting time trend of the terminal;

step 1.2, determining the reporting time t of the terminal_n；

Step 1.3, establishing a neural network of an input layer and an output layer, inputting the reporting time of each terminal into the input layer, and outputting a prediction period value by the output layer after iterative processing of the neural network;

step 1.4, comparing the size of the predicted period value of the input layer after passing through the neural network with the size of the real period value to obtain an error;

step 1.5, executing step 1.4; until the error between the size of the prediction period value and the size of the real period value of the input layer is smaller than a preset threshold value;

step 1.6, the terminals are divided into n groups.

The foregoing method for intelligent packet adaptive power saving suitable for multiple terminal systems, further, the criterion for determining whether the configuration needs to be optimized in step two includes: judging whether the terminal belongs to n groups of terminal types which are already grouped; if the terminal types are within the n types of terminal types, working according to a preset default value; if the terminal type is not in the n terminal types, entering a self-adaptive period adjustment link on the basis of the determined default value.

The foregoing method for intelligent packet adaptive energy saving suitable for multiple terminal systems, further, in step three, the DRX + eDRX adaptive adjustment loop adjustment adopts a "two-stage multi-gear" adjustment step, where, during DRX adjustment, the cycle length of the mth gear is less than the cycle length of the mth +1 gear, and m is greater than or equal to 1 and less than or equal to m_max-1, wherein m_maxThe number of the steps of the DRX period; the method mainly comprises the following steps:

step 3.1.1, if the current cycle is in the mth DRX cycle, judging whether the number of the data packets received in the current cycle reaches the threshold value N or not at the end of the cycle₁(ii) a If not, entering step 3.1.2; if so, entering step 3.1.3;

step 3.1.2, entering the (m + 1) th DRX period, and judging whether the number of the data packets in the current period reaches the threshold value N or not at the end of the period₁(ii) a If not, entering step 3.1.4; if yes, entering step 3.1.3;

step 3.1.3, judging whether the number of the data packets in the current period reaches a threshold value N₂(ii) a If not, determining the working period as T_i(ii) a If it reachesThreshold value N₂If the current m is not 1, the current cycle is set to be the m-1 th cycle, and the step 3.1.1 is returned to start execution; if the threshold value N is reached₂If m is 1, then go to step 3.1.5;

step 3.1.4, judging whether the last gear of the DRX period is reached, and entering the first gear of the eDRX period if the last gear of the DRX period is reached; if the DRX period of the last gear is not reached, returning to the step 3.1.2;

step 3.1.5, the time length corresponding to each gear period of the stage is reduced by a times, if (1-a) T₁Greater than the minimum value T of the preset period_dminThen returning to step 3.1.1 to start execution; otherwise, the adaptive adjustment loop is skipped.

The foregoing method for adaptive energy saving for intelligent packets applicable to multiple terminal systems, further, in step three, the DRX + eDRX adaptive adjustment loop adjustment adopts a "two-stage multi-gear" adjustment step, where, during eDRX adjustment, the cycle length of the nth gear is less than the cycle length of the (n + 1) th gear, and n is greater than or equal to 1 and less than or equal to n_max-1, wherein n_maxThe number of the eDRX period is the gear number; corresponding thereto, T_nmaxThe length of the last eDRX cycle. The method mainly comprises the following steps:

step 3.2.1, if the current cycle is in the nth gear eDRX cycle, judging whether the number of the data packets received in the current cycle reaches the threshold value N or not at the end of the cycle₁(ii) a If not, entering step 3.2.2; if so, entering step 3.2.3;

step 3.2.2, entering the (N + 1) th gear eDRX period, and judging whether the number of the data packets in the current period reaches the threshold value N or not at the end of the period₁(ii) a If not, entering step 3.2.4; if yes, entering step 3.2.3;

step 3.2.3, judging whether the number of the data packets in the current period reaches the threshold value N₂(ii) a If not, determining the working period as the current period T'_n(ii) a If the threshold value N is reached₂If the current n is not 1, the current cycle is set to be the n-1 th cycle, and the step 3.2.1 is returned to start to execute; if the threshold value N is reached₂Meanwhile, if the current n is 1, entering the last DRX period;

step 3.2.4, judging whether the last gear of the eDRX period is reached, if the last gear of the eDRX period is reached, entering step 3.2.5; if the last gear eDRX period is not reached, returning to the step 3.2.2;

step 3.2.5, increasing the time length corresponding to each shift period by b times, if (1+ b) × T_nmaxLess than a preset maximum period T_emaxThen returning to step 3.2.1 to start execution; otherwise, the adaptive adjustment loop is skipped.

The foregoing intelligent packet adaptive energy-saving method suitable for multiple terminal systems, further, in step three, for a period T with i paging occasions_iThere are i paging states and i sleep states, for a total of 2i states;

the states associated with the adaptive DRX + eDRX power saving policy include:

S₀: a DRX inactive state at the end of an active state, which is a critical state connecting the active state and an idle state;

S_ii ∈ N: the paging state of the DRX represents the paging time in the DRX period;

S_2i: sleep state of DRX, representing each paging occasion S_2i-1The sleep time immediately following;

S_j': the paging state of the eDRX represents the paging time in an eDRX period;

S_2j': sleep state of eDRX, representing each paging occasion S_2j-1' immediately following sleep time;

when a data packet is reported in a paging state in a DRX period, immediately returning to an active state;

when the sleep state in the DRX period has the data packet reported, returning to the active state in the next paging state;

when a data packet is reported in a paging state in an eDRX period, judging whether paging is urgent; if the paging is urgent, returning to the active state; if the paging is not urgent, the system returns to the last DRX period and continues to work.

The foregoing intelligent packet adaptive energy saving method suitable for multiple terminal systems, further, the energy saving statistics step in step four includes:

step 4.1, after the system works for a period of time, counting the power consumption of the system within a period of working time, wherein the range of the working time is set in advance;

and 4.2, if the power consumption of the system is higher than a preset threshold value, returning to execute the third step.

Compared with the prior art, the invention adopting the technical scheme has the following technical effects:

the invention enables each terminal to work in respective working period, and can reduce the pressure generated by reporting the channel by all terminals at the same time. Can freely switch between the DRX period and the eDRX period, ensure sufficient rest time when the system is idle, and ensure timely response when the system is busy.

The invention also applies DRX and eDRX simultaneously to the self-adaptive adjustment of the period, and by using the DRX and the eDRX in a matching way, the DRX period can be entered when the number of data packets is relatively large, and the eDRX period can be entered when the number of the data packets is relatively small, thereby fully utilizing the characteristics of the DRX and the eDRX. In particular, as long as the terminal can enter eDRX to operate, more energy consumption is saved than when the terminal operates by using DRX alone.

Drawings

FIG. 1 is a schematic view of the overall flow chart of the present invention;

FIG. 2 is a schematic diagram of a two-layer neural network intelligent packet of the present invention;

FIG. 3 is a diagram illustrating the transition of each state during adaptive adjustment according to the present invention;

FIG. 4 is a diagram illustrating the distribution of states in DRX and eDRX of the present invention;

FIG. 5 shows a first DRX cycle T of the present invention_iA detailed state transition process diagram;

FIG. 6 shows a first eDRX cycle T of the present invention_i' detailed state transition process diagram.

Detailed Description

The technical scheme of the invention is further explained in detail by combining the attached drawings:

it will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

The working principle of the method according to the invention is first briefly described. The energy-saving method of the invention firstly carries out intelligent grouping on the terminal, determines and configures the period default value, and then the terminal adjusts the working period and optimizes the configuration through self-adaptive adjustment loop adjustment, thereby selecting the most suitable DRX and/or eDRX period. After the terminal works for a period of time, energy-saving statistics is carried out, and whether the effect of reducing power consumption is achieved is judged. In addition, whether a new terminal exists is also judged. Therefore, on the premise of convenient management, the terminal can adjust the working period in a self-adaptive manner, thereby achieving the purpose of reducing energy consumption. On the other hand, DRX and eDRX are applied simultaneously for adaptive adjustment of the cycle, because DRX and eDRX have respective advantages and disadvantages. The single DRX period adjustment can respond to paging in time due to more paging moments, but resources can be wasted; pure eDRX cycle adjustment may have sufficient rest time due to the long sleep times at the end of each cycle, but may not allow timely response to paging. By using DRX and eDRX in a matched mode, the DRX period can be entered when the number of data packets is relatively large, and the eDRX period can be entered when the number of the data packets is relatively small, so that the characteristics of the DRX period and the eDRX period are fully utilized. In particular, as long as the terminal can enter eDRX to operate, more energy consumption is saved than when the terminal operates by using DRX alone.

Example one

The embodiment is an intelligent terminal system with a certain reporting time rule. Fig. 1 is a schematic diagram of an overall flow chart in this embodiment, and in step 100, for a system including multiple intelligent terminals, terminals are intelligently grouped first, and a default value is configured for each group. When a plurality of terminals exist in the system and are not easy to manage, the terminals can be intelligently grouped through machine learning, and a period default value is configured, wherein the period default value is a discontinuous receiving period. The preliminarily determined default value of the discontinuous reception cycle may be used as a working cycle, or an approximate range may be determined for the subsequent working cycle adjustment, so as to provide a basis for setting the cycle in the subsequent adaptive adjustment loop.

When a new terminal is added, the terminals are automatically grouped by constructing a neural network and adopting a machine learning method; and configuring a default value of the work period according to the grouping condition. And in the later stage, whether a new terminal exists is judged, whether a machine learning link needs to be returned again is determined, and grouping is carried out again.

The intelligent grouping process based on neural network is shown in fig. 2. In fig. 2, the reporting time of each terminal is preprocessed first; the reporting time of each terminal may be irregular in a short time, so that the terminal works for a period of time first, and then the reporting condition is counted. When the reporting trend tends to be stable, determining the reporting time of the terminal, namely t in fig. 2₁，t₂，t_m，t_nAnd the like. Then, by establishing two layers of neural networks, namely an input layer and an output layer in fig. 2, the reporting time of each terminal is subjected to iterative processing; and continuously comparing the error between the size of the predicted period value of the input layer after passing through the neural network and the size of the real period value. And according to the error, reversely transmitting and comparing, thereby continuously refreshing the weight and the bias among all the connections.

Furthermore, the learning process based on the neural network is not endless, since the terminal can adjust the duty cycle of the terminal by adaptively adjusting the loop at a later time, an error threshold can be determined here. And stopping grouping when the error between the size of the predicted period value and the size of the real period value of the input layer is smaller than a certain threshold value after passing through the neural network. In this example, the total division into g₁，…，g_jGroups are i groups.

After performing intelligent grouping of terminals and configuring a period default value, according to the grouping result, step 200 is performed to determine whether optimization is required. In the embodiment, whether the terminal belongs to n groups of terminal types which are grouped is judged; if the terminal types are within the n types of terminal types, working according to a preset default value; if not, then on the basis of the determined default value, the step 300 of adaptive period adjustment is entered. And determining an optimized DRX and/or an extended DRX (eDRX) working cycle through continuous self-adaptive adjustment, thereby achieving the purpose of energy saving.

For the terminal needing optimization, step 300 is executed next: and adaptively adjusting the DRX or eDRX period, optimizing the configuration and determining the work period. The adaptive adjustment mechanisms of all groups of terminals are the same, and the specific related cycle size and the threshold range are different. The whole adaptive adjustment process may go through several gear DRX cycles and several gear eDRX cycles, i.e. the adjustment mechanism of "two-stage n gear" is realized. The regulation period of the two-stage n gear comprises:

from the time perspective, the DRX periods are sequentially increased from front to back during adjustment, namely the DRX period is set as a first gear at the beginning, the period value is minimum, and the energy-saving effect is poor; continuing to increase the DRX cycle to the second gear if necessary, the cycle value being greater than the first gear cycle value, and continuing this operation if necessary until the maximum gear DRX cycle is reached. At this time, if a certain condition is met, the second level, namely an eDRX cycle adjusting stage, can be entered, and the method for adaptively adjusting the eDRX cycle is similar to DRX. It should be noted here that the maximum gear DRX cycle, although smaller than the first gear eDRX cycle, cannot be too large in difference between the two in order to achieve a perfect transition of DRX cycle and eDRX cycle.

Because of the "two-level n gear" adjustment mechanism, there are n gear DRX cycles and n gear eDRX cycles. Each period is considered as a whole and is denoted by T, and the specific adaptive adjustment process is obtained as shown in fig. 3. Wherein, one period T includes a plurality of paging occasions and sleep occasions. Next, state transition in a single period T is described.

The distribution of the states of DRX and eDRX including the paging occasion and the sleep occasion is shown in fig. 4. In FIG. 4, state S_2i-1Indicating the paging state, state S, within a DRX cycle_2iRepresenting a sleep state within the DRX cycle. State S_2j-1' indicates the paging state, State S, within an eDRX cycle_2j' denotes a sleep state within an eDRX cycle. As can be seen from fig. 4, the main difference between DRX and eDRX is in the last sleep state part during the same length cycle. For DRX, all sleep times are equally long; for eDRX, the last sleep time is much longer than the previous sleep time. Therefore, operating in eDRX mode may save more power consumption for DRX and eDRX of the same length.

First period T in DRX_iThe detailed state transition process between the respective states is shown in fig. 5. When DRX inactive state period T_inactIf a packet arrives, it stays in the state S₀ Step 501; when DRX inactive state period T_inactIf no data packet arrives, the idle state is entered, step 502. First entering the first paging state of the idle state, state S₁. Assume paging length T_poWhen T is_poIf the data packet arrives, the S is returned₀Status, step 503; when T is_poIf no data packet arrives, the first sleep state S is entered₂I.e., step 504. When in sleep state S₂If no data packet arrives, the next paging state is entered, S₃The state continues to work; when in sleep state S₂When a data packet arrives, the next paging state is just entered, i.e. S₃In the state, immediately returns to S₀The status is step 503. The state transition process in the following DRX period and the first DRX period T_iSimilarly, no further description is provided herein.

With the first DRX cycle T in eDRX_i' for example, a detailed state transition process between the respective states is shown in fig. 6. In this embodiment, the active state S₀And each state in a DRX period in an eDRX period is represented as T_kI.e. the last DRX cycle. First cycle T in eDRX_i', i.e. immediately following T_kThe state of (1). The conversion process of each state comprises the following steps:

slave state S₁' Initially, assume a paging length of T_po', when T_po' No data packet arrives in, then enter the first sleep state S₂', step 601; when T is_poIf a packet arrives, it is determined whether paging is urgent, step 602. If the paging is urgent, returning to S₀Status, step 603; if not, entering the last DRX period T_kOperation continues, step 604. When in sleep state S₂' No data packet arrives, then enter the next paging state, S₃' State continues to work; when in sleep state S₂' if a packet arrives, then the next paging state is just entered, i.e. S₃In the' state, a determination is made whether paging is urgent, step 602. The state transition process in the following eDRX period and the first eDRX period T_i' similarly.

With knowledge of the state transitions during a single DRX cycle and a single eDRX cycle as discussed above, the adaptive adjustment loop adjustments can be better analyzed. In this embodiment, the adaptive adjustment loop adjustment over multiple cycles is described in terms of three gear DRX cycle and three gear eDRX cycle, because three gear is enough to indicate the transition law between n gears.

The meaning of the parameters in fig. 3 is as follows:

T_inact: DRX inactivity period cycle length;

S₀: DRX inactive state at active state end;

T_i: representing a first gear DRX period; involving a plurality of paging states and sleep states, i.e. S in FIG. 5₁,S₂,…,S_2iStatus.

T_j: representing a second gear DRX period; same period T_iAlso includes multiple paging states and sleep states, only duration and T_iAre not identical.

T_k: representing a third DRX period; wherein, T_i<T_j<T_k。

T_i': representing a first gear eDRX period; involving a plurality of paging states and sleep states, i.e. S in FIG. 6₁′,S₂′,…,S_2jThe' state.

T_j': representing a second gear eDRX period;

T_k': representing a third gear eDRX period; wherein, T_i'<T_j'<T_k'。

N₁: representing the minimum threshold value, namely judging whether the cycle length needs to be increased continuously;

N₂: indicating a maximum threshold, i.e. the basis for determining whether a further reduction of the cycle length is required.

In this embodiment, the adaptive period adjustment mainly performs period adjustment for the idle state. At the back end of the active state, there is a DRX inactive state. If the time is within the DRX Inactivity Timer, T is_inactIf the data packet arrives, keeping the DRX inactive state all the time; when T is_inactIf no data packet arrives, the state enters an idle state. The adaptive adjustment of the idle state period is the focus of the present invention. At the end of each cycle, the total number of packets in the cycle is counted and compared with a threshold, involving two thresholds N₁And N₂The threshold range is [ N ]₁，N₂]. If the receiving number of the data packets in a period is within the range, continuing to work in the period; when the number of the data packets is less than N₁Then, a larger period is entered next time; when the number of the data packets is more than N₂Then a smaller cycle is entered next.

As shown in FIG. 3, in any DRX cycle, T_i，T_j，T_kWhen data packet arrives at any paging time, immediately returning to S₀A state; when a data packet arrives at the sleep time, returning to S at the next paging time₀Status, return to S₀The steps of the state are all indicated as step 304.

With a first DRX cycle T_iFor example, at T_iEnd of, will T_iThe number of data packets received in the receiver and the threshold N₁Step 301. If the total number of the data packets does not reach the threshold value N₁Then a longer period is entered, step 305; if the total number of the data packets reaches the threshold value N₁Then continue toJudging whether the threshold value N is reached₂Step 302. If the threshold value N is not reached₂Then the duty cycle is determined to be T_i(ii) a If the threshold value N is reached₂The previous shorter cycle is entered.

It should be noted here that, for the non-first-gear DRX cycle, when the threshold N is reached by the determination₂Then the previous shorter cycle is entered. Such as when in period T_kWhen it is judged that the threshold value N is reached₂Then enter the previous shorter period T_jI.e., step 307. When in the first gear DRX cycle, there is no preceding shorter cycle. When the final judgment exceeds the threshold value N₂Then, considering the adjustment parameters, the whole period of the stage is reduced by a times, step 308, and each period is changed to (1-a) × T_i，(1-a)*T_j，(1-a)*T_kAnd the like. Meanwhile, the minimum DRX period is ensured to be larger than the maximum value T of the preset period_dmax. Next, the adaptive adjustment loop is re-entered, step 309. The adaptive loop adjustment is re-entered in step 309, and thereafter, the original period T is entered, but each period (1-a) × T after step 308.

In any eDRX cycle, when a packet arrives at any paging time, it is determined whether paging is urgent, step 303. If urgent, go back to S₀Status, step 313; if not, the last DRX cycle is entered to operate, step 312. When a data packet arrives at the sleep time, whether paging is urgent is judged at the next paging time, and then S is determined to enter₀The state is still to enter the last DRX period for working.

At each eDRX period T_i' end, also subject to T_i' number of packets received within and threshold N₁And N₂The comparison is made, the procedure is the same as the DRX cycle. It should be noted here that when the eDRX cycle is in the last gear, the number of packets in the eDRX cycle is not N₁In this case, the readjustment parameter is considered, and the default value set initially is considered to be smaller. Here, increasing all cycles of the stage by a factor of b, step 310, is performed to change each cycle to (1+ b)*T_i′，(1+b)*T_j′，(1+b) *T_k' and the like. Meanwhile, the maximum gear eDRX period is required to be ensured to be less than the maximum value T of the preset period_emax. Next, by entering the period (1+ b) × T_k' the adaptive adjustment loop adjustment is re-entered, step 311.

In this step, the adaptive period adjustment process continuously adjusts to a longer or shorter period to optimize the working period. Generally, as long as the default value is appropriate, the terminal can work for a long time in the determined working period through self-adaptive adjustment after the working period is determined, and a better energy-saving effect is achieved.

After a period of time, step 400 is performed next: and (5) energy conservation statistics. Since the objective of the present application is to save energy, the power consumption during this period needs to be counted in case of satisfying the packet reception.

Adjusting by self-adaptive adjustment, after the working period is determined, working for a period of time (which can be days or months), and then carrying out energy-saving effect statistics. When the self-adaptive adjustment is determined to work as a DRX or eDRX period, effective statistics on energy-saving effects is difficult to be performed in a short time, and the statistical results influence the trend of the next energy-saving strategy.

And judging whether the parameters need to be reconfigured according to the energy-saving statistic result, namely step 500. If the parameters do not need to be reconfigured, continuing to execute the current operation; if a certain amount of data packets are lost in the period of time or the energy-saving effect is not obvious, the adaptive adjustment loop adjustment is considered to be re-entered, and configuration parameters are optimized. Generally, after a terminal is installed and a DRX and/or eDRX cycle is determined through learning and adaptive adjustment, the terminal may work with this configuration for a long time.

After the parameter configuration is completed, it is continuously checked whether a new terminal is added, step 600. If yes, intelligent classification is carried out, and default values are configured. The purpose of this is to make the system automatically add, delete and maintain the terminal according to the actual situation in the process of operation.

Generally, the adaptive period adjustment method described in the present application can be applied to an intelligent terminal having a certain working time rule, such as a cell having a plurality of intelligent electric meters, intelligent water meters, intelligent gas meters, and the like. For various terminal conditions controlling the same cell, the terminals in the cell are intelligently grouped by machine learning, and then different period default values are configured. And then, according to a specific working state, self-adaptive adjustment loop adjustment can be carried out to optimize configuration. All terminals select proper working cycles to work according to actual conditions, and report the working cycles separately at different time periods. And in the later stage, the addition of the terminals can be automatically carried out, then intelligent grouping is carried out, and default values are configured. Thus, the operation is repeated in a cycle. Each terminal works in the respective working period, so that the pressure generated by reporting the channel at the same time by all the terminals can be reduced. Can freely switch between the DRX period and the eDRX period, ensure sufficient rest time when the system is idle, and ensure timely response when the system is busy. Therefore, the power consumption can be reduced to the maximum extent by matching.

Different reporting requirements are provided for different terminals, and different reporting requirements are provided for terminals at different positions. Generally, the reporting period of the terminal may be in units of days, weeks, and months. For a frequently used terminal, the active time of the terminal may be set to be slightly longer; the terminal in the remote area may have a longer reporting period and may set the sleep time to be slightly longer. And setting different reporting requirements aiming at terminals with different requirements. Under the condition that the default value of the reporting period is determined, the working mode of the terminal can be set.

It should be noted that, in the method proposed by the present invention, for convenience of description, it is represented as a series of flowcharts. It will be appreciated by those skilled in the art that the present application is not limited by the illustrated ordering of acts, as various steps are executed in successive loops in accordance with the invention. Secondly, it should be appreciated by those skilled in the art that different models of different levels may be derived by applying the concepts of the present invention.

The foregoing is only a partial embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims

1. An intelligent grouping adaptive energy-saving method applicable to a plurality of terminal systems, which is characterized by comprising the following steps:

2. The intelligent packet adaptive power-saving method for multiple terminal systems according to claim 1, wherein the step of intelligently grouping terminals comprises:

step 1.2, determining the reporting time t of the terminal_n；

Step 1.3, establishing a neural network of an input layer and an output layer, inputting the reporting time of each terminal into the input layer, and outputting a prediction period value by the output layer after iterative processing of the neural network; step 1.4, comparing the size of the predicted period value of the input layer after passing through the neural network with the size of the real period value to obtain an error;

step 1.6, the terminals are divided into n groups.

3. The intelligent packet adaptive energy-saving method applicable to multiple terminal systems according to claim 1, wherein the criterion for determining whether the optimal configuration is required in step two comprises: judging whether the terminal belongs to n groups of terminal types which are already grouped; if the terminal types are within the n types of terminal types, working according to a preset default value; if the terminal type is not in the n terminal types, entering a self-adaptive period adjustment link on the basis of the determined default value.

4. The method as claimed in claim 1, wherein the DRX + eDRX adaptive adjustment loop in step three employs a two-stage multi-stage adjustment procedure, wherein the mth stage cycle length is smaller than the mth +1 stage cycle length, and 1 ≦ m for DRX adjustment_max-1, wherein m_maxThe number of the steps of the DRX period; the method comprises the following steps:

step 3.1.3, judging whether the number of the data packets in the current period reaches a threshold value N₂(ii) a If not, determining the working period as T_i(ii) a If the threshold value N is reached₂If the current m is not 1, the current cycle is set to be the m-1 th cycle, and the step 3.1.1 is returned to start execution; if the threshold value N is reached₂While at the same time presentIf m is 1, entering step 3.1.5;

5. The method as claimed in claim 4, wherein the DRX + eDRX adaptive adjustment loop in step three employs a two-stage multi-gear adjustment step, wherein when eDRX is adjusted, the nth gear cycle length is less than the nth +1 gear cycle length, and 1 ≦ n_max-1, wherein n_maxThe number of the eDRX period is the gear number; corresponding thereto, T_nmaxThe length of the last gear eDRX period; the method mainly comprises the following steps:

6. The intelligent packet adaptive power-saving method for multiple terminal systems according to claim 1, wherein in step three, for a period T having i paging occasions_iThere are i paging states and i sleep states, for a total of 2i states;

the states associated with the adaptive DRX + eDRX power saving policy include:

7. The intelligent packet adaptive energy-saving method applicable to multiple terminal systems according to claim 1, wherein the energy-saving statistic step in step four comprises: