Patent Application
20070230413
The present invention relates to communications and, more particularly, to wireless communication systems.
Wireless, radio frequency communication systems enable people to communicate with one another over long distances without having to access landline-connected devices such as conventional telephones. While early systems were primarily configured for voice communications, technological improvements have enabled the development of “3-G” (third generation) and similar wireless networks for both voice and high-speed packet data transfer. For example, CDMA-based, “1x-EVDO” (Evolution Data Optimized, or Evolution Data Only) wireless communication networks, now implemented in many parts of the U.S. and elsewhere, use the CDMA2000® 3-G mobile telecommunications protocol/specification for the high-speed wireless transmission of both voice and non-voice data. 1x-EVDO is an implementation of CDMA2000® that supports high data rates, specifically, forward link data rates up to 3.1 Mbit/s, and reverse link rates up to 1.8 Mbit/s in a radio channel dedicated to carrying high-speed packet data, e.g., a 1.25 MHz-bandwidth radio channel separate from the radio channel for carrying voice data.
Transmission latency or delay is a primary measure of system performance in wireless networks and other data systems. In a wireless network where data is transferred in packets, for example, transmission latency is typically measured as the time it takes for one or more packets to travel from a radio access network input, e.g., a radio network controller pre-buffer, to a base station including transmission out over the air. In “best efforts” 1x-EVDO wireless networks, “blocking”-type measures are unavailable, and data packets cannot be blocked based on delay. Eventually, all data bits will make it through for transmission over the air. Accordingly, one measure of how well the network is functioning (and whether it is at capacity) is the latency in transferring data packets across the wireless network.
Especially in 3-G and similar wireless networks involving high-speed data transfer, user perceived throughput (“UPT”) is also a valuable performance indicator. UPT is a sense or measure of how fast data is being received at a wireless unit, as actually perceived by the user. For example, in downloading a large file from the Internet, or in browsing web pages, users are typically not concerned with the average data throughput of the network, peak data rates, or similar generalized performance indicators. Instead, users are more typically interested in the amount of time for a batch data transfer, e.g., the time to download a particular file. In such a case, UPT may be measured as the ratio of download size (e.g., number of bits) to time, as measured from when the file is requested to when the download is completed. More generally, UPT may be characterized as the ratio of data size to the time between when the data is requested and when the data is received, in units of kbits/second (“kbps”) or the like.
An embodiment of the present invention relates to a method for measuring data transmission latency in a wireless communication network. The wireless network may include a radio access network (“RAN”) portion having one or more base stations for radio communications with various wireless units and a RAN “front end” (possibly including a radio network controller) connected to the base stations. The RAN front end acts as the interface between the base stations and the rest of the wireless network or elsewhere, including coordinating the transfer of data to and from the base stations. By “wireless unit,” it is meant mobile phones, wireless PDA's, computerized vehicle navigation systems, wireless devices with high-speed data transfer capabilities, such as those compliant with “3-G” or “4-G” standards, “WiFi”-equipped computer terminals, or the like. In carrying out the method, the transfer rate of a first data set associated with the wireless unit is determined. By “associated with,” it is meant that the data set is configured or otherwise intended for transmission to the wireless unit. For example, the first data set may be one or more data packets addressed to the wireless unit. Subsequently, the transmission latency of a second data set associated with the wireless unit is determined by applying the transfer rate to the second data set. The second data set may be, e.g., one or more data packets remaining for transmission to the wireless unit.
In another embodiment, a message is transmitted from the base station to the RNC or vice versa. The message identifies the first data set, the second data set, or a portion thereof, e.g., one of the data packets in the first or second data set. The message also includes time information, such as when the data packet was received, or when it was transmitted. The transfer rate and/or the transmission latency is determined based on the time information in the message. For example, the transfer rate may be determined as an amount of data in the first data set, divided by the time difference between the time information in the message (e.g., identifying when the last data packet in the first data set was transmitted by the base station) and a time reference point of the first data set (e.g., the time when the first data packet in the first data set was received by the RNC).
In another embodiment, a time difference is determined between first and second time reference points. The reference points and the time difference relate to a first portion of a data set. For example, the time difference may indicate the amount of time required for the first portion of the data set to pass through a section of the wireless network. The data set is associated with a wireless unit, e.g., the data set may be a plurality of data packets addressed to the wireless unit, with the first portion comprising one or more of the data packets. One of the time reference points is included in a message received from the base station or RNC and that identifies the first portion of the data set. (The first portion can be identified generally, or through the identification of a sub-component thereof, e.g., one of the data packets in the data set.) Subsequently, the data transfer rate of the first portion of the data set is determined as the ratio of the quantity of data in the first portion to the time difference.
In another embodiment, a base station in wireless communication with one of the wireless units sends successive flow control messages to the radio network controller or other portion of the RAN front end. Each message relates to the transfer a designated amount of data over the RAN airlink to the wireless unit. For example, the flow control messages may be sent each time the base station completes the transfer of the designated amount of data, for purposes of requesting additional data for transfer or otherwise. (By “transfer,” it is meant the reception and/or transmission of data, and “airlink” refers to one or more radio frequency channels or other radio links of the wireless network.) Subsequently, the current data transfer rate of the airlink, as relating to the individual wireless unit, is determined based on the designated amount of data, and on a time difference between the flow control messages. For example, the data transfer rate may be determined as the ratio of the designated data amount to the time difference or vice versa. The flow control messages may include time data such as a time stamp for determining the time difference between the flow control messages.
In another embodiment, the transmission latency of the airlink is determined based on the data transfer rate as applied to any data remaining for transfer to or from the wireless unit. For example, the data transfer rate may be calculated as the ratio of the time between successive flow control messages to the designated data amount, with the transmission latency being estimated as the product of this ratio and the amount of data remaining for transfer. Thus, for example, if the transfer of 50 data packets is completed between the sending of two successive flow control messages spaced 40 milliseconds apart, and there are 180 packets remaining for transfer to or from the wireless unit, the transmission latency would be estimated as (40 ms/50 packets) 180 packets=144 ms, based on the particular observational interval between the successive flow control messages.
In another embodiment, a user perceived throughput (“UPT”) of the airlink is determined as the data transfer rate in terms of bits per second, or as the amount of data remaining for transfer to the wireless unit divided by the transmission latency. For example, if the transmission latency were 144 ms for 180 packets, the UPT would be 1.28 Mbps (assuming 1024 bits/packet). The UPT may be calculated at a number of different times for a number of different wireless units.
Data transmission rates, transmission latency, and UPT will typically be determined with respect to individual wireless units, on a wireless unit-by-wireless unit basis. This may be done for every wireless unit in communication with the wireless network, or for only a portion of the wireless units, or at certain times, e.g., when transferring large files. Calculations made at different times for an individual wireless unit may be statistically analyzed, as may calculations across multiple wireless units.
In addition to their use in determining transmission latency, the flow control messages may be used for requesting additional data from the network front end. For example, if a base station sends a flow control message to the network front end upon completing the transfer of a designated amount of data, the flow control message may act as a prompt for the network front end to commence sending additional data to the base station. In regards to this function, the flow control messages may be an existing feature of the wireless network.
The present invention will be better understood from reading the following description of non-limiting embodiments, with reference to the attached drawings, wherein below:
With reference to
One embodiment of the method for measuring the airlink transmission latency, illustrated with respect to a base station 16a in communication with one of the wireless units 18a, is shown in
As indicated above, the method of the present invention may be implemented on different types of wireless communications networks, for example the CDMA-based 1x-EVDO network 10 shown in
The network 10 may be connected to external networks such as a public switched telephone network, or to the Internet 34. For high-speed data transmission to and from the Internet or elsewhere (e.g., for facilitating web browsing, real time file transfer, or downloading large data files), the network 10 may use the Internet Protocol, where data is broken into a plurality of addressed data packets. Additionally, voice over IP (“VoIP”) may be used for voice-data transmission. (With VoIP, analog audio signals are captured, digitized, and broken into packets like non-voice data.) Both voice and non-voice data packets are transmitted and routed over the wireless network, where they are received and reassembled by the wireless units to which the data packets are addressed. For use in transferring packet data between the RAN 12 and external networks such as the Internet 34 (or otherwise), the core IP network portion 14 of the wireless network 10 may include a packet data serving node (“PDSN”) 36 for routing wireless unit originated or terminated packet data, an authentication, authorization, and accounting module (“AAA”) 38, and a firewall 40.
In the wireless network 10, as shown in
An embodiment of the method for determining airlink transmission latency according to the present invention will now be explained in further detail with reference to
As noted above, the flow control messages 24a, 24b may each contain a time stamp or other time data of when they were generated and/or sent. Alternatively, time calculations may be based upon time data of when the flow control messages are received by the RNC 22 or otherwise. In either case, the difference “Δt” in time data (“tmessage1” and “tmessage2”) between two successive flow control messages 24a, 24b received from a particular base station 16a in regards to a particular wireless unit 18a is determined at Step 108 in
Δt=tmessage2−tmessage1
As should be apparent, Δt corresponds to the amount of time it took the base station 16a to get the designated amount of data 26 (e.g., 50 packets) out over the airlink. At Step 110, the transmission rate (“TR”) of the designated data amount 26 may be calculated as the ratio of the time Δt between successive flow control messages 24a, 24b to the designated data amount 26:
TR=Δt/(designated data amount)
To determine the transmission latency “TL,” the transmission rate TR between successive flow control messages 24a, 24b is multiplied by the data remaining for transfer 32, as at Step 112:
TL=TR·(data remaining for transfer)
TL=(Δt/(designated data amount))·(data remaining for transfer)
Conceptually, the transmission latency is the amount of time it will take for the remaining data 32 to be sent out over the airlink. Alternatively, the TL can be thought of as the time in the RAN for a new packet arriving at the pre-buffer to be sent over the airlink presuming that conditions remain “quasi-stationary.”
As an example, say that the designated data amount 26 is fifty data packets, that the time stamp on a first flow control message 24a (in regards to a particular wireless unit) indicates a time of 13:44:32.000, and that the time stamp on a second, successive flow control message 24b indicates a time of 13:44:32.040. Say also that 180 data packet exist throughout the RAN 12 for transfer to the particular wireless unit, e.g., there are 180 data packets stored in the pre-buffer 42 or elsewhere. Based on the above, the time difference Δt between the two flow control messages is 40 milliseconds:
Δt=tmessage2−tmessage1=13:44:32.040−13:44:32:000=40 ms
Then, the transfer rate TR is calculated as:
TR=Δt/(designated data amount)=40 ms/50 data packets=0.8 ms per data packet
Finally, the transmission rate is applied to the data remaining to transfer to determine the transmission latency, i.e., an estimate of how long it will take for the remaining data to go out over the air:
TL=TR·(data remaining for transfer)=(0.8 ms/data packet)·(180 data packets)=144 ms.
Thus, if conditions in the wireless network 10 do not change, the last packet of the remaining 180 data packets would take 144 ms to be sent out over the airlink. Alternatively, a new packet arriving at that instance would take 144 ms before it is at the head of the queue to be sent out over the airlink.
As noted, transmission latency will typically be determined for wireless units on an individual basis. Transmission latency may be determined for all the wireless units 18a-18c in communication with the wireless network, or only for some portion thereof, possibly based on certain types of activity. For example, determinations of transmission latency may be more relevant for situations involving large data transfers or the like. As indicated at Step 114, latency may be reevaluated periodically to capture changing RF and network conditions. Additionally, statistics can be evaluated as desired, including per-user averages, deviations, and averages over all users, using standard methods.
User perceived throughput (“UPT”) can be calculated in a similar manner as set forth above for determining transmission latency. In particular, as between successive flow control messages, UPT is determined as the ratio of bits transferred to Δt:
UPT=(designated data amount)·(bits/packet)/Δt [units: bits/sec]
(This assumes that the designated data amount is in units of packets; the designated data amount could be expressed in terms of bits, in which case there would be no need for a bits per packet conversion.) Thus, if there is a time difference Δt of 40 ms between successive flow control messages for a designated amount of data of 50 data packets, with each packet having 1024 bits:
UPT=(50 packets)·(1024 bits/packet)/40 ms=1.28 Mbits/sec (over this observation interval)
Similarly, UPT can also be calculated as the data remaining for transfer 32 divided by the transmission latency TL (again, as estimated based on a particular observation interval):
UPT=(data remaining for transfer)·(bits/packet)/TL
From the above example:
UPT=(180 packets)·(1024 bits/packet)/144 ms=1.28 Mbits/sec.
One difference to be noted is that while it is acceptable to aggregate data over all the different wireless devices (18a-18b) served by a specific BS (e.g., 16a) for the purpose of calculating the latency, for UPT the calculation has to be for a specific wireless device (e.g., for 18a only, or 18b only) as the concept of UPT is related to the throughput that is perceived or noticed an individual user.
As should be appreciated, a delay anywhere in the wireless network 10 (e.g., due to a busy transmission line or otherwise) will directly impact transmission latency and user perceived throughput. This is because an additional delay in the designated amount of data arriving at a base station will show up as an increase in the time difference At between two successive flow control messages 24a, 24b. For example (with reference to the example above where two successive flow control messages 24a, 24b are spaced 40 ms apart for a designated data amount of 50 packets), suppose that one of the transmission lines between the RNC 22 and a base station 16a becomes congested, resulting in an additional delay of 60 ms. If the next time that a flow control message is sent to the RNC 22 is 100 ms later (40 ms original delay+60 ms additional delay), the 50 packets (the designated data amount) took a total of 100 ms to transfer. If 180 data packets remain for transfer, the transmission latency would be estimated as:
TL=(Δt/(designated data amount))·(data remaining for transfer)=(100 ms/50 packets)·(180 packets)=360 ms
UPT=(designated data amount)·(bits/packet)/Δt=(50 packets)·(1024 bits/packet)/100 ms≈500 kbps
UPT=(data remaining for transfer)·(bits/packet)/TL=(180 packets)·(1024 bits/packet)/360 ms≈500 kbps
Although these examples are based on a value of 1024 bits per packet (a typical maximum value), the actual number of bits per packet may be smaller or larger than this amount. Information about the actual number of bits per packet, for purposes of calculating UPT, transmission latency, or the like, may be incorporated into the flow control messages 24a, 24b or otherwise provided in software or hardware, e.g., as a data portion of a script or computer program for carrying out the method of the present invention.
The impact of connection drops or gaps in the airlink connection would similarly be reflected in transmission latency and UPT. Also, an increased number of users would have a similar impact. In particular, with more users time-sharing the radio channel, each would have a smaller fraction of allocated slots. Thus, more time would be required for transmitting the same amount of data, which would be reflected in transmission latency and UPT. For example, suppose a first wireless unit 18a is located in a network cell or sector such that it gets a channel or slot data rate (i.e., the airlink may be logically divided into slots for transferring packet data) of 1.3 Mbps. If the user of the wireless unit downloads a file (e.g., a webpage from the Internet), and if there are two other wireless units 18b, 18c in that sector also active over the airlink, the RNC 22 (or the BS 16) will assign about ⅓ of the slots to each wireless unit 18a-18c. Thus, for the short period of several hundreds of milliseconds while the page is being downloaded, the first wireless unit 18a will perceive an effective rate of UPT=1.3 Mbps/3=433 kbps. From the perspective of the activity over the airlink as a whole, the data throughput is 1.3 Mbps. However, what is of interest to the first user is the transmission latency or UPT, as relating to his or her particular wireless unit, during times when actually transferring data, here about 433 kbps.
From the determinations of transmission latency and/or UPT as described above, jitter may also be calculated. “Jitter” is the variation in the user perceived throughput. For example, circuit-switched voice systems have a dedicated circuit with a constant UPT, meaning no jitter. In wireless networks, on the other hand, there may be variations in UPT due to factors such as variable loading (e.g., a changing number of active users) and changes in airlink conditions. Jitter may be calculated on a per-user basis, as a system average, or the like. Jitter may be calculated simply by taking the difference between successive UPT measurements, or by averaging the differences across several UPT measurements, including possibly correlating the resulting value(s) to a standardized scale or the like. For example:
UPTinterval1=500 kbps
UPTinverval2=600 kbps
Jitter=UPTinterval2−UPTinterval1=100 kbps
Determinations of data transfer performance indicators such as UPT, transmission latency, and jitter may be made at different locations in the RAN 12, depending on the configuration of the wireless network and on how the UPT and transmission latency values are to be used. For example, instead of the UPT being measured at the RNC 22, the base stations 16a-16c could be configured to make note of the time data in successive flow control messages, and to calculate the UPT based on the time data and advanced knowledge of the designated data size 26 (assuming information regarding the amount of data 32 remaining for transfer was available to the base stations).
The methods described herein for determining transmission latency, UPT, and the like may be implemented using standard hardware and/or software techniques on a wireless network's existing equipment/infrastructure. For example, the RNC 22 could be outfitted with one or more scripts (i.e., computer programs) for calculating the time difference between successive flow control messages 24a, 24b, for calculating the data transfer rate, for determining transmission latency, etc. Of course, such scripts would also be configured for transmitting the information to a designated site or component for further use. For example, as at Step 116 in
Although the method of the present invention has been primarily described in regards to forward link transmissions, it is also applicable to reverse link transmissions. For example, in transferring a file across the wireless network 10, information regarding any data remaining for transfer (e.g., file size) could be supplied by the wireless unit transferring the data, and flow control messages could be sent either (i) from the RNC 22 to the base station for requesting additional data from the base station, or (ii) from the base station to the RNC 22, at the start or completion of transferring the designated amount of data 26, as a notice that data is being transferred (i.e., instead of as a request for additional data).
As indicated, the data transfer rate, transmission latency, UPT, etc. are typically determined in part based on the time difference between successive flow control messages 24a, 24b. As should be appreciated, by “successive” it is meant any two flow control messages relating to a single data transfer event for an wireless unit, and not necessarily two flow control messages that come one right after the other. For example, it is possible that as between three temporally contiguous flow control messages, a time difference between the first and third could be calculated, provided it is known that there was an intervening message for determining that there were two “groups” of the designated data amount 26 transferred during that time period.
As should be appreciated, the method of the present invention is applicable for use in calculating transmission latency using means other than flow control messages 24a, 24b.
Latency (seconds)=remaining data/rate=M+N (packets)/rate (packets/sec)
The rate can be determined in a number of different ways, e.g., as described above with respect to flow control messages. Alternatively, the rate can be approximated at the base station 60 by measuring the how fast packets come into the base station and/or how fast packets leave the base station. For example, if a data packet number “X” is at the top of the base station queue (for transmission over the forward link) at time t=t1, and data packet number “Y” is at the top of the base station queue some later time t=t2, then the rate could be approximated as rate=# packets/time=(Y−X)/(t2−t1). (This assumes that packets are consecutively numbered and that Y>X.) Thus, if a first packet is at the top of the base station queue at time t1=0, and the fifty-first data packet is at the top of the base station queue some time later, at t2=100 ms, then the transmission rate could be approximated as TR=(51'1) packets/100 ms=0.5 packet/ms.
The rate may also be calculated through the use of messages sent from the base station to the RNC or vice versa. Such messages may also be used for determining the total number of packets for calculating latency. Generally, each message will contain (i) information identifying a data packet, and (ii) time information associated with that data packet, e.g., a time of reception, transmission, or the like. The time information in the message is then compared to a time reference point relating to that data packet or another data packet, e.g., an earlier or later time point of when that data packet or another data packet was at a particular location in the RAN 12. In effect, the transmission rate is a calculation of the amount of packets flowing past a specific point in unit time. More accurate results may be obtained by configuring the system to determine the amount of time required for a plurality of data packets to traverse the RAN 12.
For example, the RNC 22 may be configured to send messages to the base station 60 relating to the times when particular packets were sent out over the intermediate network 62. This is shown graphically in
The base station 60 receives the messages 64a-64c, and also tracks the times when the message packets arrive at the base station. For example, as indicated in
TR=total packets/(Treceive100−Tsend1)
As should be appreciated, this encompasses the entire time between when the RNC transmitted the first designated packet (packet #1) and the base station received (or transmitted over the forward link) the last designated packet, packet #100. “Designated” packet refers to packets within an observation window, not necessarily the first or last packets addressed to a wireless unit. Also, although “Tsend1” is contained in the message received at the base station from the RNC while “Treceive100” is determined at the base station, the base station and RNC have synchronized clocks for carrying CDMA communications.
As shown in
Messages between the base station and RNC may also be used for determining the total number of data packets or other data queued in the radio access network 12 for transmission to a wireless unit, for latency calculations. The messages may be used simultaneously for both rate and latency calculations. This is shown graphically in
TR=50 packets/(T3−T1) seconds
The message 68 also identifies the “earliest” packet still left in the RAN 12, here, packet #50. Since the RNC 22 knows the identity of the packet that it most recently received prior to time T3, here, packet #175, the total number of packets remaining in the RAN 12 (e.g., addressed to a particular wireless unit) is 175−50=125. The latency can then be calculated as:
Latency=remaining packets/rate=(125/50)·(T3−T1) seconds
The timing and content of the messages may vary. For example, messages may be sent based on time or on the number of received packets, as noted above. Messages do not have to be sent continually. Instead, it is possible for messages to be sent in a staggered periodic manner, e.g., messages are sent during a 1-minute period, then no messages for 4 minutes, then messages for 1 minute, and so on. Also, the measurement points may vary, e.g., time information can relate to when packets are received, transmitted, or otherwise.
Since certain changes may be made in the above-described method for measuring airlink transmission latency in a 1x-EVDO wireless network, without departing from the spirit and scope of the invention herein involved, it is intended that all of the subject matter of the above description or shown in the accompanying drawings shall be interpreted merely as examples illustrating the inventive concept herein and shall not be construed as limiting the invention.
