Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/0413—MIMO systems
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/0491—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas using two or more sectors, i.e. sector diversity
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B17/00—Monitoring; Testing
  • H04B17/30—Monitoring; Testing of propagation channels
  • H04B17/309—Measuring or estimating channel quality parameters
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
  • H04B7/0602—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using antenna switching
  • H04B7/0608—Antenna selection according to transmission parameters
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/08—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station
  • H04B7/0802—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection
  • H04B7/0805—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection with single receiver and antenna switching
  • H04B7/0814—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection with single receiver and antenna switching based on current reception conditions, e.g. switching to different antenna when signal level is below threshold
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/08—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station
  • H04B7/0802—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection
  • H04B7/0817—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection with multiple receivers and antenna path selection
  • H04B7/082—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using antenna selection with multiple receivers and antenna path selection selecting best antenna path
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04W—WIRELESS COMMUNICATION NETWORKS
  • H04W24/00—Supervisory, monitoring or testing arrangements
  • H04W24/08—Testing, supervising or monitoring using real traffic

Definitions

• MIMO Multiple Input Multiple Output
• directional antennas which use predetermined narrow beams to focus RF energy toward desired receivers. This achieves throughput gains and reduces interference. Moreover, the directivity of antennas facilitates determining proper orientation of the antenna, notably in outdoor environments allowing for a line-of-sight path. Recently, it has been shown that directional antennas in indoor environments provide a few strong paths between nodes even in absence of a line-of-sight path. The combination of MIMO and directional antennas has been discussed by N.
• the invention is based on the finding that, depending on which sectors of an antenna are activated, throughput may increase or drop when compared to using only omni-directional radiating antennas, and that increased or dropped throughput are found only in a small subset of activation patterns. Activation pattern in this
• the high-performing antenna sector patterns are determined in accordance with an average rank metric determined over a predetermined time and at different times of a day and/or days of the week.
• the inventive method generally provides a higher average throughput over any chosen time period when compared to existing MIMO wireless access points, while reducing the time needed for probing in case of a connection falling below predetermined performance thresholds.
• the inventive method may, depending on actual configuration, reduce interference with other devices operating in the vicinity through selection of appropriate antenna pattern or antenna configurations, increasing spatial reuse.
• simply combining MIMO and directional antennas not necessarily produces only benefits. For example, consider a point-to-point MIMO in an outdoor environment. Ml MO achieves capacity gains in rich scattering multi-path environments.
• an antenna selection unit is adapted to selectively activate different antenna elements of an antenna system for creating different directional antenna patterns, for transmission, Tx, and/or for reception, Rx.
• the expression 'link antenna pattern' refers to the Tx-Rx combination of directional antenna pattern at the transmitter and receiver node, respectively, of a wireless link between two MIMO wireless nodes.
• a set of link antenna pattern is determined that maximize link throughput, out of which link antenna pattern a selection is made in case the link performance drops below a predetermined threshold level.
• Link performance may be determined in terms of measured throughput, signal-to-noise ratio, or SNR, received signal strength, or RSS, packet error rate, or PER, and the like.
• one or more of the link performance metrics are determined for each one of the link antenna patterns of the set.
• one or more of the link antenna pattern from the set are selected and the link performance measurement is performed.
• the link performance measurement is done for a certain time period to obtain an average result.
• the pattern are assigned a rank in accordance with their average results, and are selected according to their rank, i.e. the pattern showing the best average rank is selected.
• sectors in each of the multiple antennas are selectively activated in such a way that successively a multiplicity of, or even all, possible combinations of activated sectors for all antennas, or link antenna patterns, are established, including activation of multiple sectors on one or more of the antennas.
• This step successively establishes a multiplicity of directional radiation beam forms for transmitting, and directional reception sensitivity pattern for receiving, for each antenna.
• the transmission and the reception pattern of each individual device may be different. For each pattern a different diversity environment is generated, which may exhibit signal paths that are advantageous in accordance with the MIMO principle.
• the different antennas may be fed with different shares of the total transmission power, further varying the signal propagation on the various signal paths.
• a received signal strength, throughput, and/or error rate, or, more general, link performance is determined.
• Error rate may include bit error rate, packet error rate, block error rate, and the like.
• the determination may be done by sending predefined data packets and receiving, from the receiver, corresponding information about the data received, or in any other known manner.
• the information so obtained is stored in a memory and the link antenna pattern are ranked in accordance with the stability of link performance over time and the respective throughput achieved. Details of the ranking will be discussed further down in this specification.
• the measurement may be repeated one or several times, immediately or after a longer period of time, in order to establish information about the temporal stability of a specific antenna combination.
• the measurement may also be repeated at different times of a day, or at different days of the week, in order to be able to provide maximum performance under varying environmental conditions that, in a regular manner, depend on the time of the day or the day of the week.
• combinations of activated antenna sectors, or sector patterns, that provide the a throughput above a predetermined threshold are combined into sets of sector patterns, the sets providing at least one combination at any time during use that provides a throughput better than a predetermined threshold.
• the wireless access point may select from the predetermined sets of sector patterns, even at short intervals, e.g. seconds, in order to maintain the best possible throughput at any time.
• a set may be valid for a certain time of the day, or for a certain day of the week.
• an average link performance is determined at predetermined time instants.
• the average link performance can be, for example, a moving average over a predefined number of past instantaneous link performance values, or an exponential average.
• the link antenna patterns are sorted in accordance with the average link performance to obtain instantaneous ranks for each of the predetermined time instants. Then the average rank of each link antenna pattern is determined for each of the predetermined time instants.
• the average can be a moving average over a predetermined number of past
• the set size can be reduced once a stable set of link antenna patterns consistently having good average ranks is found.
• the time intervals, at which the probing occurs, as well as the duration of the probing itself can be dynamically adapted depending upon how frequently the link antenna pattern with the best average rank changes. For example, in indoor systems, the pattern with the best link performance typically changes in a small time scale, following channel characteristics in typical indoor environments with people moving around. Thus, to most accurately track the channel variation, the time interval should be significantly shorter than the coherence time, which is typically around 1 second. On the other hand, one may choose a few or tens of seconds, willing to accept less accurate tracking, but amortizing the probing overheads over time.
• FIG.1 shows exemplary radiation patterns of a multi-sector antenna without feeding loss
• Fig. 4 shows exemplary STBC SN R and SDM effective SNR to throughput
• Fig. 5 exemplarily shows the throughput gain for various combinations of active sectors
• Fig. 8 exemplarily shows throughput variation over time for a 3 Tx antenna pattern set
• Fig. 9 exemplarily illustrates average RSS d i ff values for various antenna patterns
• Fig. 10 exemplarily depicts average RSS d i ff with antenna directivity gain
• Fig. 1 1 exemplarily shows the throughput gains for different numbers of link antenna
• Fig. 12 exemplarily shows throughput gains over time with and without using ranking in accordance with the present invention. Detailed description of preferred embodiments
• An exemplary apparatus in accordance with the invention is provided with at least two multi-sector antennas, each having four antenna elements.
• the antenna elements provide no directivity gain.
• the apparatus further has a processor, a program memory adapted to store program information during execution of program information, and a data memory adapted to store data during execution of the program information.
• the apparatus may further have a non-volatile storage memory, e.g. of the flash-memory type or electric erasable programmable ROM (EEPROM) type, for non- transitory storage of program information and/or data used during executing the program information.
• EEPROM electric erasable programmable ROM
• Other types of non-volatile storage memory are also conceivable, including optical or magnetic storage, and client/server data storage.
• the apparatus may further be equipped with a wired interface for receiving and transmitting data, e.g. for connecting to a LAN or to a host device.
• the wired interface may be, for example, of the USB type, or of the I EEE 802.3 type. Other types of wired interfaces are also conceivable.
• Directivity gain is used for indicating the additional antenna gain of a directional antenna toward one direction compared to an omni-directional radiation pattern.
• the antenna elements are printed on a printed circuit board, or PCB, covering the entire horizontal plane in the 5 GHz band of IEEE 802.1 1 n. Any combination of sectors can be activated for transmission or reception through a feeding network, the combinations referred to hereinafter as sector activation pattern or simply sector pattern.
• Fig. 1 (a) and 1 (b) depict the radiation patterns with one and four activated sectors, respectively.
• Table I shows that the antenna directivity gain for each pattern depends on both number of activated sectors and, for the case of two activated sectors, on whether they are opposite (2 opp) or adjacent (2 adj).
• the antenna feeding network has been designed to introduce a feeding loss such that all antenna patterns exhibit approximately equal peak gains. This design decision was made to reduce directional terminal problems in multi-hop mesh networks.
• An exemplary test setup is deployed in a single floor in a typical laboratory space, as shown in Fig. 2. This is a typical office environment consisting of cubicles, booths and offices separated by glass walls. Due to the availability of multi-sector antennas, only four nodes are needed. Different topologies can be emulated by transmission power, or Tx power, control. The exemplary measurement results presented in this specification have been obtained using UDP throughput as link performance metric. In order to measure the maximum throughput of a Ml MO link with multi-sector antennas, the large number of activation patterns and possible temporal variations need to be taken into account.
• the hardware used in the exemplary setup stores SNR information for each received packet. If a packet is encoded in the SDM mode, a pair of SNR values (SNR per spatial stream) is available, and, if encoded in the STBC mode, a single SNR value is available. In an exemplary series of measurements hardware specific SNR information is mapped to UDP throughput.
• the measurement results for different activation patterns may be affected by time variation of wireless channels.
• the fraction of SDM effective SNR samples which fall into ⁇ dB range of the long term average is measured.
• Fig. 3 shows these results for different ⁇ .
• figure 3 (b) around 15% of the samples deviate 2 dB for time intervals of 20 seconds, which is close to the 1 dB granularity of the 802.1 1 n device used.
• Fig. 4 exemplarily shows averaged throughput samples and corresponding SNR or effective SNR values for an exemplary wireless device.
• a generalized Sigmoid function is fit to each set of throughput samples with the same MCS rate.
• the throughput of a link using a certain antenna pattern is estimated by measuring the STBC SNR value and SDM effective SNR value, mapping these two SNR values to throughputs using the mappings, and choosing the maximum of these two throughput values.
• the positive throughput gains are due to the clustered propagation of signals in the angular domain. Propagation measurements in indoor environments have shown that the angles of departure, AoD, and arrival, AoA, form correlated signal clusters. Moreover, only 2 to 4 clusters contribute to the received signals. The antenna patterns that achieve positive throughput gains, are aligned in phase with these dominant signal clusters, thus avoiding negative gains, but also misaligned with other clusters that induce signal correlations at the antenna input.
• Fig. 5 shows that most links achieve negative median throughput gain, -9.3% on average, and the minimum gain can reach as low as -100%, e.g. link 1 -3 for 2 Rx sectors per antenna. Thus, less than half antenna patterns provide positive gains and if a sector activation pattern is not carefully chosen, it may yield a high penalty on throughput performance.
• the measurement results represented in Fig. 5 clearly show that the maximum throughput gains do not depend on the number of activated sectors or whether Tx or Rx activations are used.
• Link 2-3 achieves positive maximum throughput gains for both Tx and Rx activations, positive median throughput gains for Rx activation, and RSS-based throughput gains are close to maximum throughput gains for Rx activation.
• link 5-2 achieves positive maximum throughput gains only for Tx activation, has negative median throughput gains, and RSS-based throughput gains are all close to the median throughput gains.
• antenna directivity gain can be created by reducing transmit power in omni-mode.
• the network interface card, or NIC used in the exemplary test setup supports transmit power control in 3 dB increments, which allows for compensating for feeding loss of the multisector antennas when the number of active sectors is one or two, see table 1. Since feeding loss for three active sectors is 1 .25 dB, pattern sets with 3 active sectors are not considered in the following.
• Fig. 6 depicts the average throughput gains over all links, as a function of antenna pattern orientations sorted by geographical direction of activated sectors toward the other end of the link.
• TX1 and TX3 indicate 1 Tx and 3 Tx sector activation patterns, respectively.
• RX1 and RX3 correspond to Rx activation.
• F, L, R and B indicate face, left, right and back orientation, respectively.
• F/F in TX1 (or TX3) both antennas in a two- antenna Ml MO system have active Tx sectors facing the receiver.
• the geographical relationship is not correlated with throughput gains, again regardless of number of active sectors and Rx or Tx activation.
• Fig. 7 shows the throughput gains per link with antenna directivity gain.
• the maximum and median throughput gains are 71 % and 14%, respectively.
• the median gains are positive in 1 1 pattern sets out of 80, implying that more than half of activation patterns in each pattern set are likely to provide higher throughput than omni-mode. Therefore, the emulated directivity gain increases Rx signal level represented by Y even with MIMO. This effect was previously confirmed only for single-antenna systems.
• RSS can be good metric for throughput gains if antenna directivity is present, it might still be necessary to probe the RSSs of all patterns in a pattern set. This would be expected because the other spatial criteria considered above did not show any correlation. The inventors have found, however, that, by exploiting temporal properties, it is not always necessary to probe all patterns.
• the SNR measurement for activation pattern was performed from the patterns in 1 Tx, 1 Rx, 3 Tx and 3 Rx activation pattern sets. Links 2-1 , 2-3 and 2-5 are considered without antenna directivity gain. All patterns for each set were probed, and probing was repeated 40 times, spanning 4 hours from 13:30 to 17:30. Then, the throughput was obtained using the mapping discussed further above.
• Table 2 Smallest number of sector patterns that include the pattern with the largest throughput for 90% or 95% of the time.
• the average includes the RSS d it f values of all links 1 -3 and 1 -4 obtained for all 1 Sector activations of link 1 -2.
• the RSS d i ff values decrease as the number of active sectors decreases.
• the interference over omni-mode can be reduced up to 12 dB at maximum (link 2-1 ) and 8 dB on average (link 2-5).
• sector activations reduce interference level, they may not necessarily increase throughput gain.
• Fig. 9(b) depicts RSS d it f values across all antenna patterns in descending order of throughput gains.
• the patterns are first sorted in accordance with descending throughput gains and then the RSS d i ff values with the same ranking are averaged. It can be observed that, for each number of activated sectors, the RSS d it f values are not related to the throughput gains. Especially for the highest throughput gains they remain constant. Thus, by selecting a number of activated sectors, it is possible to maximize throughput gain subject to a constant interference level, which is minimum when 1 Tx Sector activation patterns are considered. In summary, the interference level without antenna directivity gain is proportional to the number of activated sectors and has little correlation with the maximum throughput gain.
• Fig. 10 depicts the average RSS d it f over omni-mode at the neighborhood of each link when antenna directivity is present.
• the average RSS d it f is at most 3 dB, and within 7 dB range of the omni-mode.
• the average 3 dB interference reduction may be too low to disable 802.1 1 n carrier sensing and thereby increase spatial reuse.
• the interference reduction does not depend on the number of activated sectors. Although two sector activation has 2.3 dB less antenna directivity gain than single sector activation, see Table 1 , its angular coverage is twice of that of a single activated sector.
• the use of multiple antennas in MI MO systems provides more opportunities for a receiver to capture more signal paths and thus receive stronger signals than a single antenna receiver.
• the interference reduction is small and does not depend on the number of active sectors.
• Implementation (2) comes with modest spatial reuse, but provides higher throughput gains and simpler protocol design. It does not require coordination among different links for sector activation. Moreover, even RSS can be used to find a good activation pattern, providing backward compatibility with IEEE 802.1 1 ⁇ .
• inventive method includes the following steps:
• the N patterns are probed to get the metric values during 7 time interval.
• the initial step enables to start from a good initial state and shortens the time to converge to a good subset.
• the steps performed during normal operation aim to find the subset of P that will include the maximum throughput patterns as many times as possible. They evaluate and select a subset within a short time, and progressively refine the subset in terms of throughput. This is done, for example, by taking the metric average, e.g., SNR, RSS or PER of the values collected so far. However, rather than using the immediate measurement results, the metric values averaged over time are used, but the subset is eventually selected based on the time averaged ranks. In this way, the inventive method weighs the order of the patterns under investigation, not emphasizing actual values of the metrics.
• the metric average e.g., SNR, RSS or PER
• P could be any subset of the entire pattern set.
• P could be a set of patterns from Tx activation with a single active sector, which are total 16. However, it is not limited to this kind of pattern. It is to be noted that, the larger the set is, the more challenging it is to find a good subset for probing. At the same time, more throughput gain can be expected.
• the time averaged rank will be leveraged to reduce the set size, reducing the number of probes.
• the choice of 7 is dependent upon how frequently the pattern with the highest throughput in P changes. Moreover, it also depends on how accurately the channel variation is tracked over time. As discussed further above, the pattern with the highest throughput changes in a small time scale, following channel characteristics in typical indoor environments. Thus, to most accurately track the channel variation, 7 should be significantly shorter than the coherence time, which is typically around 1 second. On the other hand, one may choose a few or tens of seconds of 7, willing to accept less accurate tracking, but reducing the probing overheads over time.
• the amount of additional probing overheads should be considered when N is determined. Since the overhead amount is very specific to a physical or MAC protocol used it depends on the actual implementation. For example, the probing operation may be integrated into the actual protocol as an additional packet exchange operation or as being piggybacked on existing ordinary packet transfer schedules. After that, N may be chosen such that the throughput gain outweighs the throughput loss due to the probing overheads.
• the patterns for Tx sector activation are referred to as P.
• P The patterns for Tx sector activation are referred to as P.
• SNR and RSS for all P are collected, and ranking is performed on the collected data set for fair comparison.
• the same SNR measurement is performed for both a certain pattern and omni-mode in a back-to-back manner.
• the measurement is repeated for a sufficiently number of times, e.g. 40 times, and are repeated for various times of the day, and days of the week. Measurements are repeated for all links.
• Fig. 1 1 exemplarily shows the throughput gains for different numbers of link antenna pattern to select from.
• Two metrics are considered: SNR and RSS in Fig. 12(a) and Fig. 12(b), respectively.
• the subset is determined by collecting the SNRs for all activating patterns for Rx and Tx, determining the stability of SNR over a certain time, and rank the pattern in accordance with their long-term stability. Only those pattern are considered in case the throughput falls below a threshold value, which were found to have a long-term SNR lying above a predetermined threshold value.
• the determination of SNR may be periodically repeated in order to compensate for changes in the environment.
• the invention advantageously avoids the need for brute-force testing of all possible antenna pattern in combination with excessive packet information tracking for finding an

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Quality & Reliability (AREA)
• Physics & Mathematics (AREA)
• Electromagnetism (AREA)
• Mobile Radio Communication Systems (AREA)
• Radio Transmission System (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Publications (1)

Family

ID=47844382

Family Applications (1)

Country Status (7)

Families Citing this family (9)

Citations (1)

Family Cites Families (18)

• 2012
  • 2012-03-15 EP EP12305307.6A patent/EP2639970A1/en not_active Withdrawn
• 2013
  • 2013-02-22 TW TW102106103A patent/TWI572159B/zh active
  • 2013-03-12 WO PCT/EP2013/054994 patent/WO2013135692A1/en not_active Ceased
  • 2013-03-12 CN CN201380013654.1A patent/CN104170273B/zh active Active
  • 2013-03-12 EP EP13708489.3A patent/EP2826163B1/en active Active
  • 2013-03-12 KR KR1020147025213A patent/KR102122808B1/ko active Active
  • 2013-03-12 US US14/384,551 patent/US9374140B2/en active Active
  • 2013-03-12 JP JP2014561418A patent/JP6130864B2/ja active Active

Patent Citations (1)

Non-Patent Citations (11)

Also Published As

Similar Documents

Legal Events