1、 ATIS-0500013 ATIS Standard on - APPROACHES TO WIRELESS E9-1-1 INDOOR LOCATION PERFORMANCE TESTING ATIS is the leading technical planning and standards development organization committed to the rapid development of global, market-driven standards for the information, entertainment and communications
2、 industry. More than 250 companies actively formulate standards in ATIS 18 Committees, covering issues including: IPTV, Service Oriented Networks, Energy Efficiency, IP-Based and Wireless Technologies, Quality of Service, and Billing and Operational Support. In addition, numerous Incubators, Focus a
3、nd Exploratory Groups address emerging industry priorities including “Green”, IP Downloadable Security, Next Generation Carrier Interconnect, IPv6 and Convergence. ATIS is the North American Organizational Partner for the 3rd Generation Partnership Project (3GPP), a member and major U.S. contributor
4、 to the International Telecommunication Union (ITU) Radio and Telecommunications Sectors, and a member of the Inter-American Telecommunication Commission (CITEL). For more information, please visit . Notice of Disclaimer therefore, a recommendation needs to be made for indoor testing methodologies a
5、nd evaluations for determining when indoor testing is valid. ATIS-0500013 ii FOREWORD The Alliance for Telecommunication Industry Solutions (ATIS) serves the public through improved understanding between carriers, customers, and manufacturers. The ESM Subcommittee will establish industry standards f
6、or performing E911 Indoor Testing, addressing Issue 62. The standards associated with Issue 62 will address the testing methodologies only and will not make any recommendations associated with regulator expectations or baselines. Suggestions for improvement of this document are welcome. They should
7、be sent to the Alliance for Telecommunications Industry Solutions, ESIF Secretariat, 1200 G Street NW, Suite 500, Washington, DC 20005. ESIF was responsible for the development of this publication. A. Hastings, ESIF Chair J. Goerke, ESIF 1stVice Chair G. Pavon, ESIF 2ndVice Chair S. Barclay, ATIS Di
8、rector C. Underkoffler, ATIS Chief Editor A. Nowicki, ATIS Committee Administrator The Emergency Services dense setting) Multi-story (high rise apartments/condos brick with steel framing) B. Commercial 1-2 story (commercialmasonry/brick) Multi-story (commercialmasonry) ATIS-0500013 10 Many-story (co
9、mmercial sky scrapersteel framing with glass and other materials) not highlighted in the figure Specific pictorial examples of these scenarios are provided in Annex B for illustration. In complex cities such as described in this example the test planner my find it useful to overlay a standard morpho
10、logy map, of the type commonly used in RF engineering planning tools, to highlight areas with different building densities. Some morphology maps readily available from data vendors key their morphology categories to the percentage of ground cover with construction material, thereby presenting a usef
11、ul measure of urbanization density. An example of this type of overlay is depicted in Figure 5-4 for the same area shown in Figure 5-3. It should be emphasized, however, that there are multiple approaches to arriving at the representative morphologies to derive the indoor test scenarios; these examp
12、les and figures are only provided as illustrations of possible planning techniques. Each scenario selected by the test planner should receive adequate statistical representation in the overall sample of scenarios. Since buildings of the same type will vary in their specific design, interior space, s
13、pecific construction and surroundings, as well as their distance and relative position to surrounding cell sites, multiple buildings should be selected for each building type in the test area. It is recommended that a minimum of three buildings be included in each scenario to allow for the capture o
14、f the natural variation in any given scenario type, e.g., 2-3 story apartment buildings, across the area under test. Different cell site densities or geometries within a given environment (e.g., suburban) should be reflected in the building selection whenever possible. In each selected building, a n
15、umber of test cases are identified as applicable. The exact number of buildings to be used per scenario, however, is up to the test planner and their need to meet the specific objectives of the indoor performance testing. The more buildings that are included per scenario the more reliable the overal
16、l sample, but naturally at higher costan ever present tradeoff where engineering judgment has to be prevail. One issue that can arise in the planning process is access to certain types of buildings, in particular, some residential properties. Indoor residential scenarios, although potentially more c
17、umbersome to test, are critical to capture and represent as there is an increasing trend for exclusive reliance on wireless communications in residential settings. A simple approach that could be taken during test planning to assist with indoor testing is to take advantage of the buildings of opport
18、unity that present themselves to the organizations commissioning or executing the testing. For examples, some houses or apartments where engineers, technicians or administrative staff members of these organizations live could be included when those individuals are amenable to utilizing their propert
19、ies. Care should be taken to avoid using structures of potential convenience to some technicians that would tend to bias the results, either for better or worse, because they are part of the wireless network itself, e.g., switch buildings, cell site buildings, etc. RF coverage and location performan
20、ce in these setting are non-typical of the wireless environment at large. ATIS-0500013 11 Figure 5-4: Example of Satellite View Overlaid with Morphology Map (Note: darker colors indicate denser land use) 2009 Google A fall back approach to ensure adequate scenario representation when it is otherwise
21、 not possible due to access constraints is to include what could be considered “physically equivalent” buildings. These would be buildings with similar size, construction, density and general sky visibility as those inaccessible. For example, a hotel could be used in certain cases to simulate an apa
22、rtment building, a motel to simulate a 1-3 story apartment building, or a free standing single story business to simulate a house. Due to the added uncertainties in such simulation, it should only be attempted as a last resort. In indoor testing, a fair amount of effort would typically go into ident
23、ifying a candidate building during pre-planning and in verifying, in the field, that it is indeed accessible and useable. Hence, more than one indoor test case or test point should be planned per building. This would leverage the pre-ATIS-0500013 12 planning and ground truth determination effort aro
24、und that structure and enhance the efficiency of resource utilization. The number of indoor test cases or test points per structure would typically be in the 2-6 range, depending on the type and size of structure, and the desire to sample distinct test cases within a given scenario. This number coul
25、d vary with the difficulty in determining ground truth for each test case or test point. What matters above all is that an adequate and representative overall test sample is created to meet the stated test objectives. A set of scenario examples and test associated test cases is provided below generi
26、cally, which could be tailored to the individual needs in a given test area: Residential Scenarios and Test Cases: Single family o Outside room with windows ( 3m from window) o Lower floor away from windows (e.g., hallway) Multi-family, 2-4 stories o Outside room with windows lower floor ( 3m from w
27、indow) o Outside room with windows upper floor ( 3m from window) o Building lobby if available o Interior room not on top floor (where obtaining ground truth is not impractical and cellular signal is useable) Multi-family; multi-story o Outside room with windows lower floor ( 3m from window) o Outsi
28、de room with windows upper floor ( 3m from window) o Building lobby o Entry level hallway o Upper level hallway (same position above entry level hallway) Commercial Scenarios and Test Cases: 1-2 story commercial building o Outside room with windows ( 3m from window) o Interior office or space lower
29、floor o Interior office or space upper floor Multi-story commercial building o Building lobby o Attached parking garage if available (one floor below garage roof) o Outside office with windows lower floor ( 3m from window) o Interior office or space lower floor (visible from outside office on same f
30、loor) o Outside room with windows upper floor ( 3m from window) o Interior office or space lower floor (visible from outside office on same floor) In test point selection within each test case and scenario, such as in the above example test cases, the initial selection of the test points should avoi
31、d situations where it is likely that the wireless signal will be very unreliable or the wireless service only sporadically available. At times, these are situations where there is access to wireline telephony as the prevalent mode of communication, e.g., in an ATIS-0500013 13 underground floor of a
32、business building, for example a warehouse. Such a situation would not be representative of typical or common wireless use. In each scenario, and possibly in each test point as needed, the test planner should identify the appropriate combination of ground truth determination methods to use among tho
33、se described in Clause 7 and Annex A. These will likely vary with the scenario, but could well be common among various test points and scenarios. The number of test cases identified in a group of buildings belonging to a certain scenario will often but not always be common to all those buildings (fo
34、r example, the first floor may eventually be inaccessible in a chosen building for a given scenario). The number of test points in a given building may be equal to the number of test cases in that building or it may be of larger if more than one test point is chosen for a test case. For example, two
35、 outside first floor rooms on two opposite sides of large building may be selected to take advantage of planning for ground truth determination in and access to that building. The number of test calls per test point is a variable that depends on the statistical requirements and objectives of the ove
36、rall indoor performance testing. Statistical aggregation can be performed using a combination of multiple independent test calls placed at a given test point and independent calls at another test point belonging to the same test case (e.g., 2-3 story apartment building, first floor, outside room). I
37、t is recommended that at least three independent test calls be placed from each test point. However, caution should be exercised in this approach that in the averaging performed, test cases in radically different morphologies are not combined; for example, performance in a “2-3 story building” in a
38、rural setting is not combined with that in a “2-3 story building” situated in a dense urban environment. Other test approaches may elect to perform a larger or even much larger number of calls at each test point. In such cases care needs to be followed during statistical aggregation to avoid biasing
39、 the sample with the effect of a single test point. Due to the typically large variation in indoor performance from point to point and even at a single point over time, only a sufficiently spatially and numerically aggregated sample should be interpreted to reflect indoor performance in a certain sc
40、enario and overall. 6 CREATING PLAN FOR TEST SITES A representative sample of building types needs to be selected throughout the region where indoor location performance is to be evaluated. When creating a plan for the identified test sites, the test planner should establish the following criteria f
41、or setting the objectives and assuring data needed to perform analysis of the indoor location performance has been captured, as well as being able to establish the expectations for E9-1-1 indoor performance for the tested location. It is recommended that testing be performed at one or more test poin
42、t(s) in the building representing unique performance environments. The scenarios established in Clause 5 should be followed during planning. During the planning process, selection of the suitable candidate ground truth techniques is important. Recommended ground truth solutions are outlined in Claus
43、e 7. 6.1 Number of Test Points The number of test points chosen within a structure will depend on: Size ATIS-0500013 14 One or two test points for a house; three test points for a multi-unit or multi-block structure. Complexity One test point for uniform, open construction; three or more test points
44、 for a more complex structure. Density Fewer test points may be appropriate for surrounding buildings of uniform height. More test points may be appropriate if the surrounding structures are of varying height and density. Unique Test Objectives Additional test points may be appropriate depending on
45、test objectives and the location finding system under test. 6.2 Time Interval between Test Calls With certain network implementations, data caching occurs in some parts of the location estimation process and can impact the results of any given test. Network data caching causes problems when data sto
46、red in the network equipment during a previous location attempt is reused during a later location request. For example, Test Call #1 is made and completed. Test Call #2 is then made before the data from Test Call #1 is purged, typically as a result of insufficient time having passed since Test Call
47、#1 for the data holding time to expire. When this happens the location estimate from Test Call #1 is sent in response to a bid for location on Test Call #2. The testing process should be designed to address this problem. In addition, the test planner should understand whether the location technology
48、 being tested is subject to data sharing between fixes in the device. Multiple samples should be taken at each Ground Truth location The specific number of test calls per test point depends on the objectives of the testing; however, a sample of 30 calls may be adequate for the purposes of determinin
49、g the consistency of location performance in a certain indoor performance environment. If the network is subject to data caching, there should be sufficient intervals between test calls on the same device. Indicative performance may be obtained with smaller samples. Placement of Calls Select and place test calls where adequate cell coverage to complete a voice call is available. Testing Times Test planning should account for the amount of time it will take to set up, establish ground truth and place test calls for each of the test locations. Con
