西班牙专利ES2598169A1 Method of energy saving based on microapagados for a wireless device in a telecommunication network

专利PDF首页>>西班牙专利

专利附录

专利说明

权利要求

类似技术

同族专利

引用文献

法律状态

优先权

专利摘要:
Method of energy saving based on microapagados for a wireless device in a telecommunication network. The present invention relates to a method of saving energy for a wireless device in a telecommunication network. The method comprises: receiving at least a part of a data transmission with information of duration, address of destination and address of the transmitter of said transmission; determine the fulfillment of a set of conditions, which at least comprises determining, according to the destination address, that the transmission is not directed to the device; in case said conditions are met, determine if a sum of times exceeds a minimum shutdown time previously established according to hardware limitations of the device, where said sum of time comprises at least one counter set according to the duration information ; and in case the sum is greater than said minimum time, perform a micro-off of the device with a duration equal to the value of said sum of times. (Machine-translation by Google Translate, not legally binding)
公开号:ES2598169A1
申请号:ES201631001
申请日:2016-07-22
公开日:2017-01-25
发明作者:Arturo Azcorra Saloña；Iñaki ÚCAR MARQUÉS；Francesco Gringoli；Albert Banchs Roca；Pablo SERRANO YÁÑEZ-MINGOT
申请人:Universidad Carlos III de Madrid；Universita degli Studi di Brescia；
IPC主号:

专利说明:

5
10
fifteen
twenty
25
30
35
DESCRIPTION
Jan ^ a saving method based on micro-offs for a wireless device in a telecommunication network
TECHNICAL FIELD OF THE INVENTION
The present invention has application in the field of energy efficiency and more specifically in the methods of saving energy for wireless devices connected in a telecommunication network, such as 802.11 networks, resorting to the management of micro-offs.
BACKGROUND OF THE INVENTION
Currently, IEEE 802.11 is the de facto standard for broadband Internet access. Its evolution is constant and with the latest 802.11ac updates it is possible to reach speeds of the order of Gigabit to wireless local area networks (WLAN). However, energy efficiency remains one of its main problems due to the intrinsic CSMA mechanism, which forces the wireless cards to remain active, listening to the channel at rest.
The state of the art has faced this problem, which mainly affects the devices powered by batteries, from various perspectives and offers some solutions that try to alleviate it. One of these solutions is the Power Save (PS) mode, widely deployed on commercial wireless cards, although irregularly supported by their drivers. Through this mechanism, a station (STA) can enter the sleeping state (or in English "sleep mode") for long periods of time, prior notification to the network access point (AP), if it has nothing to transmit. Meanwhile, packages directed to this sleeping station are stored by the AP and marked on the Traffic Indication Map (TIM) that travels on each beacon. From the hardware point of view, this mechanism requires support for two modes of operation: active and sleeping. This sleeping mode is implemented in the circuitry using a secondary low-frequency clock signal, so the PS mechanism dramatically reduces the consumption of a wireless card. However, the counterpart is that, since the card passes hundreds of milliseconds off, the user experiences a significant degradation of the connection performance due to the delays that are introduced.
5
10
fifteen
twenty
25
30
35
Actually, it is difficult to exploit energy saving opportunities in a CSMA protocol without disturbing its operation, since the stations need to be listening passively all the time in case they receive a package. But as the network becomes saturated, this passive listening becomes active listening of packets directed to other stations, called "overhearing", and these transmissions represent an opportunity to save energy.
According to these two operational modes (active and sleeping) discussed above, it is obvious that the more time a card spends in sleeping mode, the more savings it produces. Some of the state of the art solutions try to anticipate the PS mode by making traffic predictions. For example, detecting short intervals of inactivity (<100 ms) and putting the interface in PS mode until the arrival of the next frame according to the prediction made. These mechanisms involve a very poor granularity (tens of milliseconds) and produce degradation of the performance due to loss of frames (due to erroneous predictions), so it is only applicable to low load scenarios.
Other solutions address the problem from a different perspective, where the card's clock frequency is reduced adaptively during periods of inactivity, and returns to maximum frequency when a frame is detected. To achieve this, it is necessary to modify the physical layer (PHY) of 802.11 to include a low clock frequency operation mechanism. However, since the frames directed to other stations continue to be received, it only has good performance in low load scenarios. Additionally, the reduction of the clock frequency to implement these mechanisms is limited and, therefore, does not produce a great gain in energy compared to the sleeping mode of the current cards.
One more way, which has been explored in the state of the art, proposes to study micro-shutdown opportunities. For this, the virtual 802.11 carrier detection mechanism is used for energy saving purposes and, basically, a station sleeps when the Network Allocation Vector (NAV) or the backoff counter has a value greater than zero. Unfortunately, temporary card restrictions or actual operation are not taken into account. In addition, shutdown during the backoff window breaks with 802.11 operation: the station must listen to the channel during each backoff time slot and, if another station accesses the channel first, the backoff counter must stop to receive the plot and resume later.
5
10
fifteen
twenty
25
30
Another approach to the problem of overhearing checks the destination MAC of an incoming frame, where the wireless card goes to sleep during the duration of that frame if it is directed to another station. An instantaneous shutdown time is assumed and that the time required to return to active mode is equivalent to a Short Interframe Space (SIFS), in addition to proposing a new frame format with a new Frame Check Sequence (FCS) intended to cover MAC header only. Thus, these solutions are not compatible with the standard and also unnecessarily introduce overload in the protocol.
As explained above, the solutions known up to now by the state of the art address the problem of energy efficiency by proposing different solutions to manage active and sleeping modes, but in an unrealistic way that either sacrifices performance and addresses to scenarios with low load, or it breaks with the operation of 802.11 and risks the loss of frames. Therefore, a solution that exploits the opportunities for microapagating of wireless cards in real, saturated environments with bursts of frames, while maintaining compatibility with 802.11 and the CSMA standard would be desirable.
DESCRIPTION OF THE INVENTION
The present invention solves the aforementioned problems by taking advantage of the overheating time inherent in the CSMA protocol to perform micro-offs of the wireless device less than one millisecond. For this, in a first aspect of the invention, a method of saving energy for a wireless device in a telecommunication network comprising the following steps is presented:
a) receiving, in said wireless device, at least a part of a data transmission, wherein said part comprises information of duration, destination direction and direction of the transmitter of said transmission;
b) determine compliance with a set of conditions, where said set at least comprises determining, according to the destination address, that the transmission is not directed to the device;
c) in case the conditions of step b) are met, determine if a sum of times exceeds a minimum shutdown time previously established according to hardware limitations of the device, where said sum of time comprises at least one fixed counter of according to the duration information;
5
10
fifteen
twenty
25
30
35
d) in case the sum is greater than said minimum time, perform a micro-shutdown of the device with a duration equal to the value of said sum of times.
Optionally, the present invention contemplates that the sum of times also includes a minimum time interval between frames previously established. For example, according to the 802.11 standard, this time is defined as an SIFS.
The present invention contemplates different moments in which to initiate a micro-shutdown. One option is to wait at the end of the reception of a first frame to process the information and determine to perform a micro-shutdown of a duration according to the steps described above.
Alternatively, a preferred embodiment of the invention contemplates, once the at least part of the data transmission comprising the data of the transmission is received, with information of duration, destination direction and direction of the transmitter, to execute steps a) - d) without waiting for the reception of the rest of the transmission to end and where the sum of times in step c) also includes an additional reception time of the rest of the transmission.
According to one of the embodiments of the invention, the step of determining that the transmission is not directed to the device comprises comparing a destination MAC address with a MAC address of the device, where the comparison is performed efficiently as the byte bytes are received. the transmission in the device, so that the first different byte causes the comparison to end.
The set of conditions, according to one of the embodiments of the invention, further comprises determining that the network is in a period of contention. Optionally, determining that the network is in a contention period can be based on determining that the network is not in a contention-free period, where a contention-free period is delimited by a beacon start frame and another final beacon frame.
The set of conditions, according to one of the embodiments of the invention, further comprises determining, according to the address of the transmitter, that the transmission comes from the same network to which the device belongs.
5
10
fifteen
twenty
25
30
35
The set of conditions, according to one of the embodiments of the invention, further comprises determining that the data received is not a "clear to send" CTS frame.
The set of conditions, according to one of the embodiments of the invention, further comprises determining if the counter set according to the duration information is less than or equal to a maximum value of 32767 seconds.
In one of the embodiments of the invention, the telecommunication network complies with the 802.11 standard and the counter in step c) is a network location counter (NAV) of the wireless device.
The minimum shutdown time, according to one of the embodiments of the invention, comprises: a minimum shutdown time that the device consumes to shut down since a shutdown instruction is sent to it; a minimum power-up time the device consumes to turn on since an ignition instruction is sent to it; and a minimum preparation time that the device consumes to transmit data.
According to one of the embodiments of the invention, once the micro-shutdown is finished, the method also comprises waiting for at least a time interval equal to an established minimum interval to verify that a channel is free, less the established minimum time interval. between frames, before receiving a new transmission. For example, according to the 802.11 standard, this is equal to a DIFS time period minus a minimum SIFS interval.
According to one of the embodiments of the invention, the wireless device is a wireless card with a microcontroller configured to perform all steps of the method.
A second aspect of the invention relates to a wireless energy-saving device in a telecommunication network, comprising a wireless card and a microprocessor configured to: receive at least a part of a data transmission, wherein said part comprises information of duration, direction of destination and direction of the transmitter of said transmission; determine compliance with a set of conditions, where said set at least comprises determining, according to the destination address, that the transmission is not directed to the device; If the above conditions are met, determine if a sum of times exceeds a minimum shutdown time previously established according to hardware limitations of the device, where
5
10
fifteen
twenty
25
30
35
said sum of time comprises at least one counter set according to the duration information; and in case the sum is greater than said minimum time, perform a micro-shutdown of the device with a duration equal to the value of said sum of times.
A final aspect of the invention relates to a computer program characterized in that it comprises program code means adapted to perform the steps of the method, when said program is executed in a general purpose processor, a digital signal processor, an FPGA, an ASIC, a microprocessor, a microcontroller, or any other form of programmable hardware.
DESCRIPTION OF THE DRAWINGS
To complement the description that is being made and in order to help a better understanding of the features of the invention, it is accompanied as an integral part of said description, a set of figures in which with an illustrative and non-limiting character, what has been represented next:
Figure 1 shows the behavior of a wireless card, according to one of the embodiments of the present invention, before an RTS / CTS transmission.
Figure 2.- shows in detail the transitions imposed by the hardware in a micro-shutdown. Figure 3.- shows the response of the commercial hardware to on and off events by means of a real measurement of the behavior outlined in Figure 2. Figure 4.- shows the aggregate and normalized activity for all the stations involved in the transmission according to a simulation based in traffic taken in a real network.
Figure 5.- shows the energy consumption in mAh according to the realization illustrated in figure 4.
Figure 6.- shows the decomposition of the activity by station simultaneously with the energy saving.
Figure 7.- shows the Ewaste energy for different combinations of p and twaste.
Figure 8 shows the applicability in 802.11a for a typical transmission (frame + response ACK) in terms of the percentage of frame sizes that last a minimum of
tsleep, min, for cases with 100 ^, 200Vs and 300Vs.
5
10
fifteen
twenty
25
30
35
DETAILED DESCRIPTION OF THE INVENTION
What is defined in this detailed description is provided to help a thorough understanding of the invention. Consequently, people who are moderately skilled in the art will recognize that variations, changes and modifications of the embodiments described herein are possible without departing from the scope of the invention. In addition, the description of functions and elements well known in the state of the art is omitted for clarity and conciseness.
Of course, the embodiments of the invention can be implemented in a wide variety of architectural platforms, protocols, devices and systems, so the designs and specific implementations presented in this document are provided solely for purposes of illustration and understanding, and never to limit aspects of the invention.
The present invention discloses a method of saving energy around the main idea of putting the interface of a wireless device, such as a wireless card, in sleeping mode during those transmissions that are directed to other stations. The proposed mechanism addresses a truly practical scenario in which the micro-shutdown opportunities of 802.11 wireless networks advantageously take advantage, where it is well known that, given the intrinsic CSMA mechanism to 802.11, stations receive all frames of their SSID or others in the same channel (even some frames of adjacent channels). Upon receiving a frame, the station checks the FCS (frame verification sequence or, in English "Frame Check Sequence") to detect errors and then discards it if it is not addressed to it. In 802.11 terminology, this situation is called over-listening (overhearing.) Since overhearing consumes the energy corresponding to the reception of a complete package that was not directed to the station, it represents a substantial source of energy inefficiency.Therefore, this consumption can be avoided by means of micro-offs leading to the wireless card to a low energy mode during the transmission of said frame. In fact, the physical header PLCP (in English "Physical Layer Convergence Procedure") carries the necessary information (shipping rate and length) to know the duration of the data unit PLCP (PSDU) service, which consists of a MAC frame or an aggregate of frames. The first 10 bytes of the MAC frame indicate which is the receiver of said frame, so a frame could be discarded soon during its reception, and the station could enter a sleeping state if the circuitry is capable of supporting turned off and on in
5
10
fifteen
twenty
25
30
35
such short periods.
The present invention determines when the micro-shutdown opportunities can be exploited and the duration indicated by the NAV counter can be used to extend the duration of the micro-shutdowns, while guaranteeing the operation of 802.11 and, therefore, no frames are lost.
802.11 implements a virtual carrier detection mechanism that reduces the need to listen to the channel. With this mechanism, a station, a card or any other wireless device configured for it, is not only capable of occupying the channel during the transmission of the frame, but can also reserve the medium for a longer duration exchange with another station. For example, this exchange may include an acknowledgment (ACK) sent by the receiver, or several consecutive frames sent in a single transmission opportunity period (TXOP). For this, MAC frames have a duration field that updates the NAV ("Network Allocator Vector"), a counter that indicates how long the channel will be occupied by an exchange initiated by the current frame. This duration field is included in the First 10 bytes of the MAC header as well, therefore, the NAV can be used to extend the duration of the microapagados and, therefore, achieve greater energy savings, but according to the present invention, the opportunities to perform the microapagados must meet certain requirements for optimal operation, which requires an analysis of the hardware capabilities of the wireless card, as well as a detailed analysis of the implications of the various design decisions.
According to different modes of operation of 802.11, the operation of the NAV has certain peculiarities. On the one hand, the "Distributed Coordination Function" DCF (Distributed Coordination Function) is the basic mechanism by which all stations employ multiple access with carrier listening and collision avoidance (CMSA / CA), with a binary exponential counter. In this scheme, the duration value offers simple protection: the NAV protects until the end of a frame (data, management) plus any additional packet (control frames). For example, the ACK in response to a transmission or a " ready to send ”CTS (clear to send) + data + ACK following an RTS (request to send) request. On the other hand, when using the "point coordination function" PCF (Point Coordination Function), the time between beacons is divided into periods of contention and periods of free contention (CP and CFP respectively).
5
10
fifteen
twenty
25
30
Access point (AP) begins the CFP by setting the duration value in the beacon to its maximum value (32,768). Then, the AP coordinates communication with the stations by sending CF-poll frames. As a consequence, according to the present invention, a card cannot use the NAV to extend a micro-off during the CFP, because it must remain ready to receive a CF-poll, but it can still sleep for the duration of each package.
On the other hand, 802.11e introduces traffic categories and a new family of access methods called "Hybrid Coordination Function", HCF (Hybrid Coordination Function), which includes the "Enhanced Distributed Channel Access" EDCA (Enhanced Distributed Channel Access) and the "HCF Controlled Channel Access" HCCA. These two methods are the versions with DCF and PCF service quality respectively. Under EDCA, there are two classes of duration values: single protection, as in DCF, and multiple protection, where the NAV protects until the end of a sequence of frames in the same TXOP.HCCA works similarly to PCF, but under HCCA, the access point starts the CFP dynamically.In the CFP, when the access point sends a CF-poll to a station, puts the NAV of other stations at a value equal to TXOP, which depends on the traffic category, however, if the station finishes transmitting before the end of TXOP, the access point can claim that reset time the NAV of the rest of the stations with another CF-poll. As a result, according to the present invention, the NAV cannot be used, either in this family of access methods, to extend a micro-shutdown during a CFP.
Finally, an additional situation that the present invention contemplates as a limitation for the use of the NAV refers to 802.11g, to interoperate with the networks 11b, the CTS-to-self frames are introduced. These are standard CTS frames, transmitted at an inherited rate and not preceded by an RTS, which are sent by a station destined for itself to occupy the channel before sending another frame. In this case, a station cannot know what the destination of the next frame will be. Therefore, wireless cards, according to the present invention, should not use the duration field of a CTS to make micropayments.
Taking into account these limitations, the present invention, to ensure that no frames are lost when introducing the proposed method of energy saving, also contemplates
5
10
fifteen
twenty
25
30
35
The problems of hidden node and capture effect, since, although they are not described by the standard, are well known and studied effects in the literature that do occur inevitably in real deployments. The design of the algorithm proposed by the present invention, according to one of its embodiments, limits the opportunities for microapagating a wireless card to the packets of the same network.
It is well known that a high power transmission can completely blind another with a lower signal to noise ratio (SNR). Theoretically, two stations occupying the channel at the same time give rise to a collision, but in reality, if the power difference is high enough, a wireless card is capable of decoding the high power frame perfectly ignoring the other transmission. Basically, there are two types of capture effect depending on the order of the frames: if the high power frame starts first, the situation is equivalent to receiving a frame and some noise to continuation, so it does not affect the behavior of a card that implement the method of the present invention. In the opposite case, the receiving card interrupts the decoding of the low power frame and goes on to decode the other with more power. If this frame arrives during the first PLCP, the method of the present invention will still not have made any decision, so it will not be affected. However, since a high power transmission can blind a low power one at any time, even when the data transmission itself (the MAC frame) has started, the scenario in which a station receives low power frames must be considered. from distant networks. If the access point is not able to see such frames, a hidden node problem arises. Obviously none of these frames is directed to our station, but if we try to save energy by making micro-shutdowns during those transmissions, potential transmissions (of greater power) from the network itself, directed to it and that if they were to be received by capture effect would be lost . Therefore, the algorithm of the present invention contemplates that, under hidden node conditions, frames that if received under normal conditions could be lost and consequently proposes to avoid micro-offs when packets come from networks other than their own. To discard packets originating from other networks, the basic service identifier BSSID is checked at the destination address for frames directed to the access point. Thus, if the frame was sent by an access point, it is only necessary to look at 6 additional bytes (in the worst case), which are included in the address of the transmitter, which does not necessarily imply more modulation time. For example, for 11ag OFDM rates, this implies an additional 8 ^ s at 6 and 9 Mbps, 4 ^ s at 12, 18 and 36 Mbps, and no additional time at 24, 48 and 54 Mbps.
5
10
fifteen
twenty
25
30
Meeting all the above limitations, a preferred embodiment of the present invention, which also adds the temporary restrictions imposed by the real hardware, saves, during the transmissions of the network itself in which the station is not involved, sleeping during the Overhearing periods or, in English, "overhearing" and taking advantage of the NAV mechanism when possible, which, as explained above, occurs during CP periods while avoiding CTS frames.
After a micro-shutdown, two things can happen:
• The card wakes up at the end of a frame exchange. For example, after data + ACK. All stations must wait for an interval of space between DIFS frames before returning to content.
• The card wakes up in the middle of a frame exchange. For example, as shown in Figure 1, a card (9) intends to establish a communication with a second card (12), for which it sends an RTS request (1) that other cards also receive (11) that are not recipients of the communication, which means setting the NAV counter (2) until the end of a fragment (3) plus the ACK (4) and the card takes the opportunity to perform a micro-shutdown (10). This first fragment fixes the NAV (5) until the end of the second fragment (6) plus the corresponding ACK (7), but the sleeping card obviously does not see it. Once the card wakes up and returns to the active state, a period of silence of duration corresponding to the minimum interval between frames, SIFS (8) is observed, and then comes the second fragment, which fixes the NAV again.
This implies that a station can remain in a sleeping state for an additional SIFS, as shown in Fig. 1, and wait in active mode for a minimum of one DIFS before returning to contention.
So far, it has been described how long a microapagado can last. Next, the limitations imposed by the hardware are described and which, according to the present invention, are fundamental to define how short a micro-shutdown can be, or, what is the same, what is the minimum period supported by a card. For this purpose, the parameters shown in Figure 2 are defined below, where the vertical axis is the power and the horizontal is the time:
toff (22): the time required to go from active mode (21) to sleep (20).
5
10
fifteen
twenty
25
30
ton (23): the time required to return asleep.
Tready (24) the time required by the circuitry to stabilize and be ready to transmit or receive packets.
From these parameters, the minimum time, tsleep, min, is calculated as:
Asleep, min = toff + ton + ^ eady (1)
The steps of the method proposed by the present invention can be clearer by observing the algorithm which, by way of example, is included below. Said algorithm describes, according to one of the embodiments of the present invention, the main loop of the microcontroller of a wireless card that implements said realization of the present invention.
Once the first 16 bytes of the frame have been received, the receiver knows the duration (tNAV), the destination direction (Ra) and the transmitter address (7a). One of the embodiments contemplates that the MAC addresses be compared efficiently as the bytes are received, so that the first different byte (if the first Ra byte has the multicast bit to zero, that is, Ra is unidifusion (unicast)) the procedure to enter sleeping mode (SET SLEEP) is launched. Additionally, the main loop maintains a global variable (C) that indicates whether the network is in contention period (CP) or not (CFP). The implementation, according to one of the embodiments of the invention, is based on detecting the beacon frames that indicate the beginning and the end of each CFP.
The procedure (called "SET SLEEP" in the example algorithm) to activate the sleeping mode and cause a micro-shutdown takes as input the remaining time until the end of the frame in the air (tDATA) and the duration value (tNAV). The latter is only used if it is a valid value and the network is also in a CP period. Then, the card performs a micro-shutdown of duration tsleep = tDATA + tNAV + an SIFS interval (as long as the total sum is greater than the minimum time of the card set above), wait for a DIFS interval to complete and return to the main loop:
5
10
fifteen
twenty
25
30
35
40
Example Algorithm:
one: ...
2: global C ^ true 3: loop
4: ...
5: while bytes remaining do
6: Read
7:
8:
9:
10:
if Ra = BSSID OR (Ta = BSSID AND Ra is other unicast MAC) then | Sleep set (£ data, £ nav)
end if end while
eleven:
FCS check
12:
13:
14:
fifteen:
16:
17:
if is Beacon AND tNAV> 0 then
| C ^ false else if is CF End then
| C ^ true end if
18 end loop
19:
twenty:
twenty-one:
22:
2. 3:
24:
procedure Set SLEEP (fDATA, tNAV)
Asleep ^ ^ DATA + ^ SIFS
if C AND is not CTS AND tNAv
I ^ sleep ^ ^ sleep + ^ NAV
end if
if ^ sleep - ^ sleep, min then
25:
26:
27:
28:
SLEEP (tsleep) WaIT ( Difs - tsiFs) go to Main loop end if
<32767 then
29:
go to Receiving loop
30: end procedure
1> Initialisation 1> Contention flag 1> Main loop
1> Receiving loop
1> Frame received 1> CFP starts
1> CFP ends
Additionally, the present invention dispenses with mechanisms of cyclic redundancy to detect errors, since the impact of errors in the first 16 bytes of the MAC header (those important for the algorithm decisions that take advantage of micro-shutdown opportunities) do not cause degradation of the performance or error rates (BER) higher than those given for normal channel conditions.
5
10
fifteen
twenty
25
30
As for the hardware part involved, as previously introduced, the present invention advantageously takes advantage of the micro-shutdown opportunities offered by an 802.11 network, taking into account some limitations that guarantee its best performance, but also, the hardware capabilities of the wireless device used are fundamental to adjust the algorithm and get its application in real scenarios, where transition times should not be neglected. The toff, ton and tready times are particular for each device, so it is necessary to analyze them experimentally to characterize it.
Next, by way of example, a particular case is described in which an Atheros AR9280 Half-height Mini PCI Express wireless card has been used, but with any other model the characterization would be carried out in the same way. In this case it is an 11abgn-capable card that supports up to 2 spatial streams and 40 MHz channels (up to 300 Mbps). Its main clock has a frequency of 44 MHz for 20 MHz channels and 88 MHz for 40 MHz channels in the 2.4 GHz band, and 40 MHz for 20 MHz and 80 MHz channels for 40 MHz channels in the 5 GHz band. This watch turns off completely and gives way to a secondary 32 kHz when transitioning to low energy mode or sleeping mode.
To proceed with the characterization, the card is connected to a computer through a flexible adapter, for example an x1 PCI Express to Amfeltec Mini PCI Express. This adapter connects the bus data channels for peripheral devices (PCI) to the computer and provides an ATX type connector for external power. The same computer contains a high precision multifunction acquisition card (DAQ), optimized for an accuracy of 18 bits. Its temporal resolution is 50 ns with an accuracy of 50 ppm. In this way, the operations sent to the card through the computer and the energy measurements made with the DAQ can be correlated using the same time base. The power supply can be, for example, a Keithley 2304A type of direct current, which is optimized to test wireless communication devices that operate with batteries. The power supply is produced through a measuring circuit that extracts the voltage and converts the current with a precision measuring resistor and an amplifier. Considering that the DAQ has some stabilization time, it is modeled as a small capacity that acts as a low pass filter. Therefore, two voltage followers are established as input to the DAQ to reduce the output impedance of the measurement circuit. Finally, with the assembly described, to obtain the minimum shutdown time (tsleep, min) supported by the
5
10
fifteen
twenty
25
30
card, the parameters toff, ton and tready defined above and represented in Figure 2 are measured.
Then, from the kernel space, the following steps are performed to characterize the card according to this experiment:
1) Initially, the card is in active mode and connected to an access point in 11a mode.
2) A pure connection (Linux AF PACKET socket) is created and a small fake package is prepared.
3) A card shutdown is forced by writing a record. This writing is virtually instantaneous in kernel time.
4) A delay of 60 ^ s is introduced to allow time for the card to react.
5) The ignition is forced by writing another register.
6) A configurable time is expected.
7) The false frame is sent at a low level, i.e., calling the function ndo_start_xmit () from net_device directly. In this way, it ensures that the execution is very fast.
The hardware response is shown in Figure 3. As can be seen, the card spends 50 ^ s consuming in active mode from when the shutdown event is launched until it is actually turned off. The power on event is launched 10 us later. And, again, the card needs another 50 ^ s to return to active mode. Note that the transmission of the packet is launched right after the power-up event and that the transmission lasts very little time in kernel (the time of these instructions corresponds to the width of the “start_xmit” rectangle). But, surprisingly, the card sends the package 200 ^ s after returning to active mode, although the plot was ready much earlier.
To understand the reasons for this delay in the transmission, more experiments are carried out in which to launch the frame with different delays, also represented in the same figure 3 for delay values between the ignition event and the transmission of the 0 packet (31 ), 200 (32) and 350 (33) ^ s, and the conclusion is that the card always starts transmitting at the same point regardless of when the kernel space transmission is launched within those first 250 ^ s after the event of ignition. Beyond that point, the transmission begins almost instantaneously. These experiments show that this specific card needs fready = 200 ^ s to be operational again.
5
10
fifteen
twenty
25
30
In short, this characterization shows that, if we want to perform a micro-shutdown with this card for a certain time / sleep, it must be taken into account that it supports a minimum / sleep time, min of 300 ys. Therefore, / sleep ^ / sleep, min must be satisfied, and the interruption must be programmed to lift the interface again 250 ys before the end of the shutdown period.
It should also be noted that the card wastes energy for a constant time / waste:
/ waste - / off + / ready (2)
equal to 250 and also. Therefore, the total time the card spends in the low energy state is / sleep - / waste.
This characterization has been made for a particular card, but the same process can be repeated for any other wireless device that interoperates in an 802.11 network and is intended to be implemented according to the present invention. Energy savings are maximized, while ensuring the operation of the network even in saturated scenarios.
To reinforce the advantages of the present invention, some evaluations of the energy savings achieved in a real network that include wireless cards such as the one characterized above and implement the method proposed by the present invention are presented below. To perform such an evaluation, 802.11a traces have been used with some 44 million packages, divided into 43 files, of the public data set SIGCOMM’08.
The methodology used for each trace file, according to the evaluation of this particular realization, consists in that, in a first pass, all the stations and access points present are discovered. Next, each station is mapped to its BSSID and a bit array is implemented to maintain the status at every moment (connected or disconnected). Since it is difficult to determine from a capture when a station is disconnected, because they almost always disappear without sending a disassociation frame, a criterion is used, usually followed by access points (AP), whereby a station is consider connected if you have transmitted at least one frame in the last 5 minutes. In a second pass, the amount of time that each station spends (without applying the algorithm of the present invention) in the states of transmission, reception, over-listening and active (considering that the stations are measured)
5
10
fifteen
twenty
25
30
connected are always active; i.e., if a station announces that it is going to PS mode, it is ignored). In a third pass, the amount of time that each station spends (now with the algorithm of the present invention) in the same states plus low energy (sleeping) is measured. It is verified that, as expected, the transmission and reception times coincide with the previous pass. As part of the active time (idle) is included the time wasted by micro-offs due to hardware limitations (twaste). After this processing, there are several duplicate identifiers (MAC addresses) (i.e., a station that appears in more than one file). Those entries are added by time for each state. At this point, the activity time is defined as the sum of transmission, reception, over-listening, sleeping and wasted time. The active time in any of these states is not taken into account because the objective is to estimate how much energy can be saved in these periods, which are the only ones that are given while the channel is occupied. Using this definition, it is evident that most stations have very little activity (a few seconds participate and disappear). Therefore, filtering is performed by the top decile to obtain the 42 stations with the highest activity.
Figure 4 represents the aggregate and normalized activity for all stations. The fraction of the activity time is represented on the vertical axis depending on the state of the card (transmission (41), reception (42), over-listening (43), low energy time (44), wasted time (45), time active spaces between frames (46)). The transmission and reception times are labeled "common" (47) because the stations use the same time in these states with (48) or without (49) the algorithm used in the present invention. It is evident that the mechanism of the present invention effectively reduces the overheating time from a median of approximately 70% to 30%. The card consistently uses less overheating time because this time difference, plus some active time of the spaces between frames, becomes micro-offs, that is, low energy time and wasted time.
This activity aggregation allows to calculate the total consumption using the data obtained in the card characterization. Figure 5 shows the energy consumption in mAh units (assuming a typical 3.7 V battery) and according to the realization described above and illustrated in Figure 4. The energy saved exceeds 1200 mAh (15.8%) even with restrictions imposed by this card, which prevent, on the one hand, that it sleeps when the time is not long enough and, on the other hand, it wastes a great deal of time setting up the circuit in each micro-shutdown.
5
10
fifteen
twenty
25
30
Figure 6 shows a breakdown by station. The graph below shows the decomposition of the activity by season, according to the algorithm of one of the embodiments of the present invention (the bars corresponding to the transmission (41), in white, are barely appreciated). Interesting is the fact that the over-listening time is reduced more or less to a constant fraction for all stations (ie, with the algorithm, this time represents more or less 30% of the total activity per station), while the Less participatory stations (left part of the graph) spend more time sleeping. The upper graph shows the consumption per station with the algorithm plus the energy saving (60) in dark gray, which is of the order of tens of mAh per station.
To demonstrate the practical implications of the present invention, it is necessary to insist that the gains of the proposed algorithm depend on the behavior of the circuitry involved. Its capabilities, in terms of temporary restrictions, determine the maximum savings that can be achieved. Particularly, each microapagado wastes (in comparison with an ideal scheme in which the wireless card passes in low energy mode all the microapagado) a fraction of energy that is given by:
Ewaste (t) = (1 ~ P) + P
Where:
p = Psleep / Poverh ~ psleep / pidle; and twaste is the fixed part defined in Equation (2).
Figure 7 represents said Ewaste energy for the particular realization described above, in which an AR9280 card (p = 0.3, twaste = 250) is used together with other values. Clearly, although a low p is always desirable (we see the evolution in the 3 curves that correspond to values p = 0.1 (72) and p = 0.3 (71)), a twaste improvement is essential to improve the performance in Short micro-offs (the evolution in the curves is reflected for the values of 50 jjs (73), 150 jjs (74) and 250 jjs (75)). Similarly, the minimum shutdown time, tsleep, min limits the applicability of the present invention. For example, considering a common case in 11ag networks, the transmission of a frame (up to 1500 bytes) followed by the corresponding ACK:
tsleep, min ^ tDATA + tSIFS + tACK + tSIFS
19
5
10
fifteen
twenty
25
30
and expanding the right side,
t, <
sleep, mm
8 (14 + I min + 4)
Tie
+ t SIFS + tPLCP +
8 (14 + 2)
^ Any
+ tr
Where Imin is the minimum amount of data (in bytes, and removing the MAC header and FCS) that a frame must contain to last the minimum sleep time feleep, min. Based on this Imin, Figure 8 defines the applicability in 802.11a in terms of the percentage of frame sizes (<1500 bytes) that last a minimum of / sleep, min, for cases with 100®s (81), 200®s (82) and 3006®s (83). Again, an improvement in / waste would drive not only the savings by micro-off but also the applicability.
On the other hand, it should also be taken into account that the applicability may be affected by the evolution of the standard. Particularly, 802.11n introduced, and 802.11ac followed, a series of changes that allow elevating speeds up to Gigabit. These improvements are based mostly on MIMO methods and the use of more coupled channels. However, a 20 MHz channel remains more or less equivalent to case 11ag. It is true that some improvements allow to improve the speed in a single channel from 54 to 72 Mbps under optimal conditions, but it is also true that the PLCP is longer to accommodate the complexity of the new coding schemes. This addition not only extends each transmission, but encourages the use of frame aggregation. Therefore, the increase in capacity, in these and future amendments, does not necessarily imply shorter channel usage times, so the solution provided by the present invention remains applicable.
In any case, in the future, reducing temporary card restrictions is essential to improve energy savings. First, / off and / on should depend on the implementation of the internal firmware (i.e., the complexity of saving and restoring the state). Secondly, the transmissions are much more aggressive, in terms of sudden change in power, than in the return from the low energy state, so the time needed to stabilize the card characterized above seems to show an obvious margin of improvement. And thirdly, the 802.3 standard goes beyond
5
10
fifteen
twenty
25
30
35
802.11 and, although in a limited way, it defines new transition parameters for the interfaces (e.g., tw phy would be equivalent to ton + tready) that are in the range of tens of ys in the worst cases. Due to all these reasons, the wireless card manufacturers carry a development line that invites us to think that in a short time the capabilities of the invention proposed here will be even enhanced and more advantageous savings schemes can be offered to the extent that Progress in standardization activities.

权利要求:
Claims (15)
[1]
5
10
fifteen
twenty
25
30
35
1. - Jan. saving method for a wireless device in a telecommunication network, characterized in that it comprises the following steps:
a) receiving, in said wireless device, at least a part of a data transmission, wherein said part comprises information of duration, destination direction and direction of the transmitter of said transmission;
b) determine compliance with a set of conditions, where said set at least comprises determining, according to the destination address, that the transmission is not directed to the device;
c) in case the conditions of step b) are met, determine if a sum of times exceeds a minimum shutdown time previously established according to hardware limitations of the device, where said sum of time comprises at least one fixed counter of according to the duration information;
d) in case the sum is greater than said minimum time, perform a micro-shutdown of the device with a duration equal to the value of said sum of times.
[2]
2. - Method according to revindication 2, where the sum of times also includes a minimum time interval between frames previously established.
[3]
3. - Method according to any of the preceding claims wherein, once the at least part of the data transmission comprising the data of the transmission is received, with duration information, destination address and transmitter address, it further comprises execute steps a) -d) without waiting for the reception of the rest of the transmission to end and where the sum of times in step c) also includes an additional reception time of the rest of the transmission
[4]
4. - Method according to any of the preceding claims, wherein the step of determining that the transmission is not directed to the device comprises comparing a destination MAC address with a device MAC address, where the comparison is performed efficiently as they are received. the bytes of the transmission on the device, so that the first different byte causes the comparison to end.
[5]
5. - Method according to any of the preceding claims wherein the set of conditions also comprises determining that the network is in a period of contention.
[6]
6. - Method according to any of the preceding claims wherein the set of conditions also comprises determining, according to the address of the transmitter, that the transmission comes from the same network to which the device belongs.
5
10
fifteen
twenty
25
30
35
[7]
7. - Method according to any of the preceding claims wherein the set of conditions also comprises determining that the data received is not a "clear to send" CTS frame.
[8]
8. - Method according to any of the preceding claims wherein the set of conditions also comprises determining if the counter set according to the duration information is less than or equal to a maximum value of 32767 seconds.
[9]
9. -Method according to any of the preceding claims, wherein the telecommunication network complies with the 802.11 standard, the counter in step c) is a network location counter (NAV) of the wireless device.
[10]
10. - Method according to any of the preceding claims, wherein the minimum shutdown time comprises: a minimum shutdown time that the device consumes to shut down since a shutdown instruction is sent to it; a minimum power-up time the device consumes to turn on since an ignition instruction is sent to it; and a minimum preparation time that the device consumes to transmit data.
[11]
11. - Method according to revindication 5 where determining that the network is in a contention period is based on determining that the network is not in a content free period, where a content free period is delimited by a frame start beacon and another final beacon plot.
[12]
12. - Method according to any of the preceding claims wherein, once the micro-shutdown is finished, it also comprises waiting at least a time interval equal to a minimum established interval to verify that a channel is free, less the established minimum interval of time between frames, before receiving a new transmission.
[13]
13. - Method according to any of the preceding claims wherein the wireless device is a wireless card with a microcontroller configured to perform all steps of the method.
[14]
14. - Wireless energy saving device in a telecommunication network, comprising a wireless card and a microprocessor configured to: receive at least a part of a data transmission, where said part comprises information on duration, direction of destination and direction of the transmitter of said transmission; determine compliance with a set of conditions, where said set at least comprises determining, according to the destination address, that the transmission is not directed to the device; in case the above conditions are met, determine if a sum of times exceeds a minimum shutdown time previously established according to hardware limitations of the device, where said sum of time comprises the
minus a counter set according to the duration information; and in case the sum is greater than said minimum time, perform a micro-shutdown of the device with a duration equal to the value of said sum of times.
[15]
15. Computer program characterized in that it comprises program code means 5 adapted to perform the steps of the method according to any of claims 1 to 13, when said program is executed in a general purpose processor, a digital signal processor, a FPGA, an ASIC, a microprocessor, a microcontroller, or any other form of programmable hardware.
10

类似技术:

公开号 | 公开日 | 专利标题

US10667169B2|2020-05-26|Supporting a large number of devices in wireless communications

ES2696831T3|2019-01-18|Procedures and devices for low power activation signal and operations for WLAN

ES2703189T3|2019-03-07|Unchained functioning in objective waking time

CN109644402B|2020-10-23|System and method for sending wake-up message

ES2617597T3|2017-06-19|Procedure for transmitting and receiving a signal from a usable station in energy-saving mode in a wireless communication system, and corresponding device

US9628243B2|2017-04-18|Communication method with indications in the PHY header

ES2676532T3|2018-07-20|Systems and procedures to reduce collisions after paging of traffic indication maps

ES2441165T3|2014-02-03|Method for the operation of a mobile terminal as an access point and corresponding mobile terminal

ES2611935T3|2017-05-11|Random access channel enhancements for LTE devices

ES2708970T3|2019-04-12|Devices for the paging of reduced overhead

US20160330788A1|2016-11-10|Mobile radio communication devices and methods for controlling a mobile radio communication device

AU2013244199B2|2017-07-13|Method and apparatus for accessing channel in WLAN system

US7716379B2|2010-05-11|Hardware control interface for IEEE standard 802.11 including transmission control interface component and a transmission status interface component

ES2763941T3|2020-06-01|Transmission and reception of discovery and paging messages

ES2665794T3|2018-04-27|Systems and procedures for transmitting configuration change messages between an access point and a station

US7689164B2|2010-03-30|Relay apparatus, communication terminal, communication system, and semiconductor integrated circuit

ES2886530T3|2021-12-20|Wireless communications method and wireless communications terminal in a high-density environment including an overlapping basic service set

EP2836046B1|2018-05-30|Method and apparatus for accessing channel in wlan system

ES2618960T3|2017-06-22|Methods and devices to indicate the transmission time of packets

ES2618956T3|2017-06-22|System and method for association with a basic service group

WO2014098367A1|2014-06-26|Backoff method and device in slot-type channel access of wireless lan system

ES2742029T3|2020-02-12|Procedure and apparatus for sending and receiving data on a machine-to-machine wireless network

ES2642097T3|2017-11-15|Procedure for resting at least one component of an entity of a corresponding communication network, device and computer program

US10073798B2|2018-09-11|Hardware control interface for IEEE standard 802.11

ES2626497T3|2017-07-25|Power management of communication devices

同族专利:

公开号 | 公开日

EP3490309A4|2020-03-11|

EP3490309A1|2019-05-29|

ES2598169B1|2017-08-29|

WO2018015601A1|2018-01-25|

引用文献:

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

US20060072614A1|2003-03-26|2006-04-06|Sony Corporation|Radio communication system|

WO2011046701A2|2009-10-18|2011-04-21|Intel Corporation|Device, system and method of selectively aborting reception of wireless communication packets|

US20130003628A1|2009-11-13|2013-01-03|France Telecom|Method for deactivating at least one component of an entity of a communications network, corresponding device and computer program|

US7277417B2|2003-04-29|2007-10-02|Broadcom Corporation|Low power protocol for wireless terminal peer-to-peer communications|

法律状态:
2017-08-29| FG2A| Definitive protection|Ref document number: 2598169 Country of ref document: ES Kind code of ref document: B1 Effective date: 20170829 |

优先权:

申请号 | 申请日 | 专利标题

ES201631001A|ES2598169B1|2016-07-22|2016-07-22|Micro-power-based energy-saving method for a wireless device in a telecommunication network|ES201631001A| ES2598169B1|2016-07-22|2016-07-22|Micro-power-based energy-saving method for a wireless device in a telecommunication network|

PCT/ES2017/070520| WO2018015601A1|2016-07-22|2017-07-18|Energy-saving method based on micro-shutdowns for a wireless device in a telecommunication network|

EP17830541.3A| EP3490309A4|2016-07-22|2017-07-18|Energy-saving method based on micro-shutdowns for a wireless device in a telecommunication network|

[返回顶部]