200525167 九、發明說明： 【發明所屬之技術領域】 本案係關於在通訊網路中藉基地台發射之通訊信號之 傳播行爲的映照和模擬。 【先前技術】 以諸如無線局部區域網路（Wireless LAN)、藍芽 (Bluetooth)、全球行動通訊系統（GSM)或泛歐式數位無線電 話系統（DECT)等技術爲基礎的無線通訊系統正被廣泛地應 用於各種場合，它們不但存在於工業生產場所及辦公環境 中，亦無所不在地存在於醫療保健場所。 藉通訊系統發射之通訊信號所產生之電磁場的傳播特 性實質上決定了該通訊系統關於涵蓋範圍（c 〇 v e r a g e)、可利 用率（availability)以及傳輸率（transmission rate)等方面的 表現會b 力（performance capability)。 一方面，無線網路操作者有興趣的是決定場傳播特性或 信號特性的分佈，例如：電磁場強度、位元錯誤率（bit error rate)、信號雜訊比（signal-to-noise ratio)等，以便能理想地 計畫該無線網路以便能夠表現出所使用之系統在網路設置 之後的品質確保或是能夠診斷出系統運作時的錯誤狀況。 另一方面，網路服務提供者有興趣的是能夠提供用戶定 址（location-dependent)服務給它們的顧客。 爲了這個目的，必須知道接收裝置的位置，只有在正常 的網路運作下所產生的資料被使用於位置估計時，觀察信號 特性在這裡才有意義。 200525167 先前技術中已提出了不少關於諸如DECT行動電話或無 線網路中配置無線局部區域網路的個人數位助理（PDA)和筆 記型電腦的定位終端（1 〇 c a t i n g t e r m i n a 1)的方法及程序。 在某些方法中（例如引用文獻[1]所提），定位 (localization)僅係以網路拓樸爲基礎，終端的位置係由其所 連接之基地台以及由其連接經歷所決定。 然而，這種方法的正確性很低，只有該終端所連接之該 基地台附近的一極大範圍內才被特定爲一可能位置。 其他已知的方法皆試圖以所有可用之傳輸器 (transmitter)所接收到的場強度來估計位置，在某些情形 下，甚至會使用應用於波傳播的詳細實體模型；然而，爲了 這個目的，便有需要獲悉與環境相關的詳細資訊。 引用文獻[2]提出了一種使用建築物中之不同阻隔 (wall)之電磁特性之知識的方法，然而，一般來說，這種知 識是無法應用的，在大部份的情形中，一場強度映照（m a p) 會因此而被首先決定，並接著被應用於定位。 接收端位置的點估計經常係基於場強度映照而被執行 (引用文獻[3]及[4])。 引用文獻[5]提出了一種應用於DECT行動電話之位置 估計的遞迴隨機非線性過濾（1^。111<3丨¥651〇(：11&81丨(：1101：11丨116&1· filtering)技術。 由先前技術中得知的大部份的這些方法皆係以手動 (manual)方式測量映照，但其在實際使用時會招致令人無法 接收的損失/複雜性。 -6- 200525167 職是之故，申請人鑑於習知技術所產生之缺失’經過悉 心試驗與硏究，並一本鍥而不捨之精神，終構思出本案「在 通訊網路中藉基地台發射之通訊信號之決定用以描述其傳 播行爲的映照之方法、配置、帶有程式碼裝置之電腦程式及 其電腦程式產品」，以下爲本案之簡要說明。 【發明內容】 本案之目的係提供一種簡單、平價且自動的程序，以產 生該種映照。 爲達到本案之目的，提供一種在通訊網路中藉基地台發 射之通訊信號之決定用以描述其傳播行爲的映照之方法、配 置、帶有程式碼裝置之電腦程式及其電腦程式產品’其具有 後述之申請專利範圍獨立項的特徵。 藉由決定該映照的方法，與所測量之相關所選定位置有 關的一物理通訊信號特性係於該通訊網路中的所選定位置 上被測量，該物理特性代表該通訊信號的該傳播行爲。 用於該傳播行爲的一模型係藉由使用所選定位置而決 定，或是藉由使用所選定位置之對應位置資訊以及該通訊信 號的所測量之相關物理特性而決定。 該映照係藉由使用用於該傳播行爲的該模型而決定’該 模型本身係已被選定成爲一映照或是該模型本身即已爲一 映照。 根據本案發明，在所選定位置上進行的測量係使用一自 主型（autonomous)行動單元而執行，例如一自主型自動控制 裝置（autonomous robot)，設置該自主型行動單元係以接收該 200525167 通訊信號及藉由一適當的測量裝置以測量該通訊信號的該 物理特性。 用以決定該映照的配置係具有一自主型行動單元，設置 該自主型行動單兀係用以接收該通訊信號及以測量該通訊 信號的該物理特性；例如藉由一適當的測量裝置。 該物理特性代表該通訊信號的該傳播行爲。 該自主型行動單元係於該通訊網路中的選定位置上進 行測量，在該通訊網路中的選定位置上，與所選定之相關位 置有關的該物理通訊信號特性係於每個情況中被測量。 此外，該配置更具有一估計單元，用於該傳播行爲的一 模型係使用該等選定位置而被該估計單元決定、或是使用該 等所選定位置之對應定位資訊以及該通訊信號之該等相關 測量物理特性而被該估計單元決定，該映照係使用用於該傳 播行爲的該模型而被該估計單元決定。 此處需要注意的是，該模型本身係已被選定成爲一映照 或是該模型本身即已爲一映照。 本案之一優點在於，該自主型行動單元能夠以一自動方 式主動進行測量，意味著該映照可以自動產生。 貝p使_於該通訊網路之運作進行中，仍可依序使得該映 照獲#更使其能適應該網路架構中的變動（例如：個別 傳輸站的失效）。 # ^ ® —種具有程式碼裝置的電腦程式，其係根據 本案之方法的所有步驟而在一電腦上執行時以決定該映照。 # ^ M tt另一種具有程式碼裝置的電腦程式產品，其係 200525167 根據本案之方法的所有步驟而在一電腦上執行該程式時，其 、 係儲存於一機器可讀取式資料媒介中，用以決定該映照。 . 本案所設置的該配置以及具有程式碼裝置的該電腦程 式’係根據本案之定位方法在一電腦上執行該程式時，施行 所有的步驟，而本案所提供的另一種電腦程式產品，係根據 本案之定位方法而施行所有步驟而在一電腦上執行該程式 時’其係儲存於一機器可讀取式資料媒介中；其皆係特別 適用於實現依據本案之方法或是下列所說明的其中一種發 展。 馨 本案之較佳實施例皆係根據本案之申請專利範圍獨立 · 項而顯示。 下述的各項發展方案皆係同時關於方法和配置以及軟 體構成。 下述的發明及各項發展方案皆可同時於軟體及硬體內 應用，例如：使用一特殊的電子電路而完成。 再者，後述之本案的發明或發展的應用係藉由一電腦可 讀取儲存媒介而完成，可執行本案發明的具有程式裝置的電 β 腦程式係儲存於該電腦可讀取儲存媒介上。 後述之本案的完成或每個發展方案係藉由一電腦程式 產品而完成，該電腦程式產品具有一儲存媒介，可完成本案 發明的具有程式裝置的電腦程式或發展方案係儲存於該儲 存媒介上。 對於一通訊網路（例如一無線網路）中介於一行動通訊 裝置（行動站；例如一行動電話）以及一基地台（例如一全方 200525167 向性的天線或是一全方向性的發射器或是一或多個不同天 線）之間的通訊來說’資料（該通訊信號）係以爲人所知之突射 (b u r s t)的信號封包方式被傳輸。 基於所傳輸的或所發射之通訊信號或突射的（可測量） 物理特性，不同之與距離相關的（distance-related)參數便可 以被決定，其係依序被用來當作決定發射特性或（信號）傳輸 器之信號特性的基礎。 這樣一種與距離相關的（亦即：取決於距離的）參數係爲 例如：一通訊信號或突射的一場強度、一位元錯誤率或是一 信號雜訊比。 一發射通訊信號的場強度很自然地取決於與一傳輸器 以及（呼叫處理）基地台之間的距離，並因此而提供關於該傳 輸器之傳播行爲（傳播特性）的資訊。 場強度、（一般以物理特性來看）以及與傳輸器或（接收） 定位或通訊網路中的（接收）位置之間的關係能夠以物理模 型來描述，該物理模型係描述信號的傳輸行爲。 在一種所謂的前向模型（forward model)中，通訊信號的 物理特性係以該通訊網路中的一位置或是一距離的函數來 描述。 一後向模型（backward model)或反向模型（illverse mo del)可將該通訊網路中的位置以通訊信號之物理特性的 一函數來描述。 與引用文獻[5 ]相似的前向模型型態的一適用模型具有 一決定性（deterministic)元件（次要模型）和一隨機 200525167 (stochastic)元件（次要模型）。 該決定性元件表示該通訊信號的物理特性以及該通訊 網路中的一位置之間的相依，該隨機元件表示該決定性元件 的一不確定性（uncertainty)。 該決定性元件的該不確定性可以是該通訊信號的一不 確定性，特別是測量雜訊、及/或該相依的一不確定性，特 別是該傳播模型的一不確定性。 大型的通訊網路一般皆具有複數個或大量的基地台，每 個基地台皆會發射一通訊信號。 建議於此處建立複數個映照或模型，在每一情況中之一 映照或模型皆代表該通訊網路中藉一基地台所傳輸之一通 訊信號的一傳播行爲。 複數個映照或模型亦可結合以形成一整體映照。 此外，建議利用該自主型行動單元以決定所選定之位置 或其定位資訊，因此，該自主型行動單元必須配置一位置測 量系統以決定該通訊網路中該自主型行動單元的該位置、及 /或必須配置一路徑設計系統以決定該通訊網路中的一移動 路徑。 已知實施例中包括了里程計及/或影像處理位置測量系 統、及航行位移計算系統。 本案提出之程序或方法而產生的該映照可以作爲通訊 網路中眾多應用的基礎。 舉例來說，該映照或該模型及/或該整體映照或該整體 模型亦可使用於設計（planning)及/或配置（in st all at ion)及/ -11- 200525167 或委任（commissioning)及/或診斷（diagn〇sing)錯誤狀況及/ 或該通訊網路的品質確保。 用於該傳播行爲的該模型及/或該映照及/或用於該傳播 行爲的該模型及/或該整體映照亦可使用於該通訊網路中以 定位至少一行動通訊裝置’其中設置的至少一行動通訊裝置 係用以接收該通訊信號及/或用以接收該等通訊信號。 在該通訊網路中所決定的一位置上，會針對該位置上可 接收到的通訊信號或其物理特性進行一測量，使用該映照或 該模型可接著自所測量的物理變數進行計算以決定該位置。 本案亦將該映照或該模型的產生以獨立於任何定位的 方式提供，該映照接著便能夠藉由該自主型行動單元及本案 方法而持續獲得更新，其中，使用已產生的映照可在該通訊 網路中定位行動通訊裝置。 本案或該映照的本案發明係特別適用於一數位蜂巢式 行動無線系統的環境中，例如：一 G S Μ網路，其可用以針 對一 G S Μ電話（行動電話）進行定位。 使用本案發明時，僅使用行動電話之可資利用的資料， 而不會針對G S Μ網路或G S Μ網路中的基地台進行高價的改 4^： 〇 本：案亦可適用於其他數位蜂巢式行動無線系統的環境 中’例如：一無線局部區域網路、一以藍芽（Bluet00th)爲基 礎的網路或一泛歐式數位無線電話系統（DECT)網路，例如其 可用以定位一 DECT行動電話。 【實施方式】 -12- 200525167 實施例：無線網路中信號特性的自動映照 實施方式如下所述： · 首先說明無線網路映照的基本原理。 其次說明無線網路之基地台的傳播特性及形成，其亦包 括該映照係如何由一決定性的（deterministic)和一隨機的 (s t 〇 c h a s t i c )部份所構成。 同時亦將說明以該映照爲基礎的定位。 接著是所提出程序之實驗證明的描述，此處將會針對用 以自動測量場強度的一自動控制系統進行說明，亦將其映照 Φ 品質納入考慮。 . 1 .映照 _ 在正常運作期間，諸如DECT行動電話的無線網路接收 器會持續測量所有有效之傳輸站的場強度。 如果已知傳輸站的精確位置，並且得出電磁波傳播的一 適切物理模型，則其資料便可用於估計使用者的位置。 在實際方案中，傳輸站的位置一般係屬於未知、或是不 夠精確的已知且易於變動。 ® 引用文獻[6]認爲，當處於一複雜的建築特徵之模式 下、特別是位於室內時，便難以或幾乎不可能產生一精確的 物理波傳播模型。 因此，以下便首先產生個別傳輸器之場強度的一映照， 藉由該映照可以定位該終端° 該場強度映照可以下列測量模型描述： -13- 200525167200525167 IX. Description of the invention: [Technical field to which the invention belongs] This case is about the reflection and simulation of the propagation behavior of the communication signal transmitted by the base station in the communication network. [Previous technology] Wireless communication systems based on technologies such as Wireless Local Area Network (Wireless LAN), Bluetooth, Global System for Mobile Communications (GSM) or Pan-European Digital Radio Telephone System (DECT) are being widely used Places are used in various occasions. They exist not only in industrial production places and office environments, but also everywhere in medical care places. The propagation characteristics of the electromagnetic field generated by the communication signal transmitted by the communication system essentially determine the performance of the communication system in terms of coverage (availability), availability, and transmission rate. (Performance capability). On the one hand, wireless network operators are interested in determining the distribution of field propagation characteristics or signal characteristics, such as electromagnetic field strength, bit error rate, signal-to-noise ratio, etc. In order to ideally plan the wireless network so as to show the quality of the system used after the network is set up or to diagnose the error condition of the system during operation. On the other hand, network service providers are interested in being able to provide user location-dependent services to their customers. For this purpose, it is necessary to know the position of the receiving device. Observing the signal characteristics is only meaningful here if the data generated under normal network operation is used for position estimation. 200525167 A number of methods and procedures have been proposed in the prior art regarding positioning terminals (10 c a t i n g t e r m i n a 1) such as personal digital assistants (PDAs) and laptops configured with wireless local area networks in DECT mobile phones or wireless networks. In some methods (for example, cited in [1]), localization is only based on the network topology, and the location of the terminal is determined by the base station to which it is connected and its connection experience. However, the accuracy of this method is very low, and only a large range near the base station to which the terminal is connected is specified as a possible location. Other known methods attempt to estimate the position from the field strengths received by all available transmitters, and in some cases, even detailed physical models applied to wave propagation are used; however, for this purpose, There is a need to learn more about the environment. The reference [2] proposes a method of using knowledge of the electromagnetic properties of different walls in a building. However, in general, this knowledge cannot be applied. In most cases, a field strength Maps are therefore determined first and then applied to positioning. Point estimation of the receiver's position is often performed based on field-strength mapping (references [3] and [4]). Citation [5] proposes a recursive random non-linear filtering (1 ^ .111 < 3 丨 ¥ 651〇 (: 11 & 81 丨 (: 1101: 11 丨 116 & 1 · 1) applied to the position estimation of DECT mobile phones). filtering technology. Most of these methods known from the prior art measure the mapping manually, but they will cause unacceptable loss / complexity in actual use. -6- 200525167 In view of the lack of knowledge technology, the applicant, after careful testing and research, and a spirit of perseverance, finally conceived the case "The decision to borrow the communication signal transmitted by the base station in the communication network is used to Describe the mapping method, configuration, computer program with code device and its computer program product ", the following is a brief description of this case. [Summary of the Invention] The purpose of this case is to provide a simple, affordable and automatic program In order to achieve this kind of mapping, in order to achieve the purpose of this case, a method of providing a mapping of communication signals transmitted by a base station in a communication network to describe its mapping behavior , Configuration, computer program with code device and its computer program product 'which has the characteristics of the independent items of the scope of patent application described later. By determining the mapping method, a physical communication related to the selected location measured The signal characteristic is measured at a selected location in the communication network, and the physical characteristic represents the propagation behavior of the communication signal. A model for the propagation behavior is determined by using the selected location or by Determined using the corresponding location information of the selected location and the measured relevant physical characteristics of the communication signal. The mapping is determined by using the model for the propagation behavior 'The model itself has been selected as a mapping or It is the model itself that is already a reflection. According to the present invention, the measurement performed at the selected location is performed using an autonomous mobile unit, such as an autonomous robot, which sets the autonomous Type mobile unit is to receive the 200525167 communication signal and to use a suitable measuring device to measure The physical characteristic of the communication signal. The configuration used to determine the mapping has an autonomous mobile unit, and the autonomous mobile unit is set to receive the communication signal and measure the physical characteristic of the communication signal; An appropriate measurement device. The physical characteristic represents the propagation behavior of the communication signal. The autonomous mobile unit measures at a selected position in the communication network, and at a selected position in the communication network, The characteristics of the physical communication signal related to the relevant position are measured in each case. In addition, the configuration has an estimation unit, and a model for the propagation behavior is determined by the estimation unit using the selected locations, Or it is determined by the estimation unit using the corresponding positioning information of the selected positions and the relevant measured physical characteristics of the communication signal, and the mapping is determined by the estimation unit using the model for the propagation behavior. It should be noted here that the model itself has been selected as a mapping or the model itself is already a mapping. One of the advantages of this case is that the autonomous mobile unit can take active measurements in an automatic way, meaning that the mapping can be generated automatically. In the operation of the communication network, it is still possible to make the mapping in order to adapt to changes in the network architecture (for example: failure of individual transmission stations). # ^ ® —A computer program with a code device that determines the mapping when executed on a computer in accordance with all the steps of the method in this case. # ^ M ttAnother computer program product with a code device, which is 200525167. When the program is executed on a computer according to all the steps of the method of this case, it is stored in a machine-readable data medium. Used to determine the mapping. The configuration set in this case and the computer program with a code device are all steps performed when the program is executed on a computer according to the positioning method of the case, while another computer program product provided in this case is based on When the positioning method of this case executes all steps and executes the program on a computer, it is stored in a machine-readable data medium; they are all particularly suitable for implementing the method according to the case or one of the following descriptions A development. Xin The preferred embodiments of this case are shown according to the independent items of the scope of patent application in this case. The following development schemes are all about method and configuration and software composition. The following inventions and various development schemes can be applied in both software and hardware, for example: using a special electronic circuit to complete. Furthermore, the application of the invention or development of the present invention to be described later is completed by a computer-readable storage medium, and the electric β-brain program with a program device that can execute the invention of the present invention is stored on the computer-readable storage medium. The completion of this case described below or each development plan is completed by a computer program product, the computer program product has a storage medium, and the computer program or development plan with a program device that can complete the invention of this case is stored on the storage medium . For a communication network (such as a wireless network) between a mobile communication device (mobile station; such as a mobile phone) and a base station (such as a omnidirectional 200525167 directional antenna or an omnidirectional transmitter or It is the communication between one or more different antennas) that the data (the communication signal) is transmitted in the form of a known burst signal packet. Different distance-related parameters can be determined based on the transmitted or transmitted communication signals or the physical properties of the burst (measurable), which are used in order to determine the transmission characteristics The basis of the signal characteristics of an or (signal) transmitter. Such a distance-dependent (i.e., distance-dependent) parameter is, for example, a field strength of a communication signal or burst, a bit error rate, or a signal-to-noise ratio. The field strength of a transmitting communication signal naturally depends on the distance from a transmitter and the (call processing) base station, and therefore provides information on the propagation behavior (propagation characteristics) of the transmitter. The field strength, (generally viewed in terms of physical characteristics), and the relationship with the transmitter or (receiving) positioning or (receiving) location in the communication network can be described by a physical model that describes the transmission behavior of the signal. In a so-called forward model, the physical characteristics of a communication signal are described as a function of position or distance in the communication network. A backward model or illverse model can describe the position in the communication network as a function of the physical characteristics of the communication signal. A suitable model of the forward model type similar to the reference [5] has a deterministic element (secondary model) and a stochastic 200525167 (stochastic) element (secondary model). The decisive element represents a physical characteristic of the communication signal and a dependency between a position in the communication network, and the random element represents an uncertainty of the decisive element. The uncertainty of the decisive element may be an uncertainty of the communication signal, especially a measurement noise and / or an interdependent uncertainty, especially an uncertainty of the propagation model. Large communication networks generally have a plurality of base stations, and each base station transmits a communication signal. It is suggested that a plurality of mappings or models be established here. In each case, one mapping or model represents a propagation behavior of a communication signal transmitted by a base station in the communication network. Multiple maps or models can also be combined to form a whole map. In addition, it is recommended to use the autonomous mobile unit to determine the selected location or its positioning information. Therefore, the autonomous mobile unit must be equipped with a position measurement system to determine the location of the autonomous mobile unit in the communication network, and / Or a path design system must be configured to determine a moving path in the communication network. The known embodiment includes an odometer and / or image processing position measurement system, and a navigation displacement calculation system. The mapping generated by the procedure or method proposed in this case can serve as the basis for many applications in a communication network. For example, the mapping or the model and / or the overall mapping or the overall model may also be used in planning and / or configuration (in st all at ion) and / -11-200525167 or commissioning and And / or diagnosing error conditions and / or quality assurance of the communication network. The model and / or the mapping used for the propagation behavior and / or the model and / or the overall mapping used for the propagation behavior may also be used in the communication network to locate at least one mobile communication device ' A mobile communication device is used to receive the communication signals and / or to receive the communication signals. At a position determined in the communication network, a measurement is performed for the communication signal or physical characteristics that can be received at the position, and the mapping or model can then be used to calculate from the measured physical variables to determine the position. This case also provides the mapping or the model generation in a manner independent of any positioning. The mapping can then be continuously updated by the autonomous mobile unit and the method of the case. The generated mapping can be used in the communication network. Locate a mobile communication device on the road. The present invention or the invention of the reflection is particularly applicable to the environment of a digital cellular mobile wireless system, such as a GSM network, which can be used to locate a GSM phone (mobile phone). When using the invention of this case, only the available data of the mobile phone will be used, and no expensive changes will be made to the base stations in the GS Μ network or the GS Μ network. 4 ^: This case can also be applied to other digital In the context of a cellular mobile wireless system, for example: a wireless local area network, a Bluetooth-based network, or a pan-European digital wireless telephone system (DECT) network, for example, it can be used to locate a DECT mobile phone. [Embodiment] -12- 200525167 Example: Automatic mapping of signal characteristics in a wireless network The implementation is as follows: First, the basic principle of wireless network mapping is explained. Secondly, the propagation characteristics and formation of the base station of the wireless network will be explained. It also includes how the mapping is composed of a deterministic and a random (s t o c h a s t i c) part. Positioning based on this mapping will also be explained. Next is a description of the experimental proof of the proposed procedure. An automatic control system for automatically measuring field strength will be described here, and its reflection Φ quality will be taken into consideration. 1. Mapping _ During normal operation, wireless network receivers, such as DECT mobile phones, continuously measure the field strength of all valid transmitting stations. If the precise location of the transmission station is known and a suitable physical model of electromagnetic wave propagation is obtained, its data can be used to estimate the user's position. In practical solutions, the location of the transmission station is generally unknown, or is not known accurately enough and is subject to change. ® Citation [6] states that it is difficult or almost impossible to produce an accurate physical wave propagation model when it is in a complex architectural feature model, especially indoors. Therefore, the following first generates a map of the field strength of an individual transmitter, with which the terminal can be located. The field strength map can be described by the following measurement model: -13- 200525167
-hl(xk)" 丄 η" 少 N(xk)- 十 .VN. hM + vk (1) 其代表了所有傳輸器的對數測量信號強度 f Pk …ιο·ι〇4』 ⑺ 其中，卜1……n係爲一決定性的非線性函數hf(Xk)，其 係取決於該接收器的位置座標Xk^XhkXhk]1'。 V k代表該對數測量的相加錯誤，其係對應於未知衰減所 造成的倍數不確定。 在無線網路中所g寸論的該傳輸器在通訊時係藉由所傳 輸的一惟一 ID來識別，因此其可進行成功的測量以精確地 指定測量方程式之一，以大大地簡化估計問題。 在正常網路運作下即使在定位（localization)階段也不 會發生干預（intervention)，因此使用者或接收裝置並不需要 額外的偵測器，定位因此必須只基於接收器之場強度測量而 產生。 2 .映照與定位 如第1圖所示，一場強度映照1 2 0首先藉由自動控制裝 置1 1 〇基於測量1 1 1和一估計參考位置1 1 2而產生於映照階 段1 〇〇，該映照1 2 0包括次要映照1 2 1，每個次要映照代表 一基地台的傳輸特性。 在後續的定位階段1 5 0中，持有一接收裝置（例如一 DECT行動電話或一無線局部區域網路或配置藍芽的pda) 的複數個（人）使用者160係基於該映照120而被定位165。 -14- 200525167 第2圖表示兩種階段的同時實施，映照階段1 〇 〇及定位 階段1 5 0。 在過程中，當使用者1 6 0已被定位1 6 5、1 7 0時，自動 控制裝置1 1〇會更新映照120。 這意味著其可永遠保證有一更新的映照1 2 0至網路的改 變，使得諸如錯誤或個別傳輸器的更換能立即被倂入。 映照階段（100) 決定性的測量模型 對於映照所奠基的決定性測量模型來說，其係假設一傳 輸器的對數接收場強度會隨著距離而線性地減少。 對於二維的位置座標xk來說，基地台^之對數場強度 的線性減少可藉由下列的測量方程式來描述 h^(xk) = - ms)T + Δς (3). 在因此獲得的N個方程式（每個基地台一個）中，總共6 N 個參數可由mf，以及〜來估計。 隨機的次要模型 對於在實際的傳輸器特性以及測量模型之間的測量雜 訊和誤差來說，必須將不確定的隨機次要模型列入考慮。 模型不確定性 模型不確定性代表了對數接收場強度的精確模型以及 決定性的測量模型in(xk)之間的差別。 第3圖顯示其差別。 該錯誤一方面係源自於該模型，然而，另一方面，其亦 -15- 200525167 代表了源自於特殊建築特性的局部誤差，其包括了不同的衰 減或不同材料的反射特性，該模型不確定性在這裡可使用中 位數和標準差σ<：^來描述。 測量雜訊 測量雜訊代表了在一點上所測量的時間變化，舉例子來 作說明，第3圖顯示測量値如何偏離於實際的傳輸特性。 測量雜訊亦可由中位數/2/和標準差〇*<2>來描述。-hl (xk) " 丄 η " Less N (xk)-X.VN. hM + vk (1) It represents the logarithmic measurement signal strength f Pk of all transmitters… ιο · ι〇4 』』 Among them, Bu 1 ... n is a deterministic non-linear function hf (Xk), which depends on the position coordinates Xk ^ XhkXhk] 1 'of the receiver. V k represents the addition error of the logarithmic measurement, which corresponds to the uncertainty of the multiples caused by the unknown attenuation. In the wireless network, the transmitter is identified by a unique ID transmitted during communication, so it can perform successful measurements to accurately specify one of the measurement equations, which greatly simplifies the estimation problem. . Under normal network operation, even during the localization phase, there is no intervention. Therefore, the user or the receiving device does not need an additional detector. Therefore, the positioning must be based on the field strength measurement of the receiver. . 2. Mapping and Positioning As shown in Figure 1, a field map 1 2 0 is first generated by the automatic control device 1 1 10 based on the measurement 1 1 1 and an estimated reference position 1 12 in the mapping phase 1 0. Map 1 2 0 includes secondary maps 1 2 1, each secondary map representing the transmission characteristics of a base station. In the subsequent positioning phase 150, a plurality of (human) users 160 holding a receiving device (such as a DECT mobile phone or a wireless local area network or a Bluetooth-equipped pda) are based on the mapping 120. Be positioned 165. -14- 200525167 Figure 2 shows the simultaneous implementation of two phases, mapping phase 1000 and positioning phase 150. In the process, when the user 160 has been positioned 16 65, 170, the automatic control device 110 will update the image 120. This means that it can always guarantee an updated image of 120 changes to the network, so that errors such as or replacement of individual transmitters can be immediately imported. Mapping phase (100) The decisive measurement model For the decisive measurement model based on mapping, it is assumed that the logarithmic receiving field strength of a transmitter decreases linearly with distance. For a two-dimensional position coordinate xk, the linear decrease in the logarithmic field strength of the base station ^ can be described by the following measurement equation h ^ (xk) =-ms) T + Δς (3). In each equation (one for each base station), a total of 6 N parameters can be estimated by mf and ~. Random secondary models For uncertain transmitter noise and measurement noise and errors between measurement models, uncertain random secondary models must be considered. Model uncertainty Model uncertainty represents the difference between an accurate model of the logarithmic receiving field strength and the decisive measurement model in (xk). Figure 3 shows the difference. This error originates from the model on the one hand, but on the other hand, it also represents a local error derived from special building characteristics, which includes different attenuation or reflection characteristics of different materials. The model Uncertainty can be described here using the median and standard deviation σ <: ^. Measurement noise Measurement noise represents the time variation measured at one point. For example, Figure 3 shows how the measurement noise deviates from the actual transmission characteristics. Measurement noise can also be described by median / 2 / and standard deviation 0 * < 2 >.
v，C 結合不確定性 整體模型的不確定性係爲模型不確定性和測量雜訊的 結合，爲了獲得一簡單模型，此處假設該不確定性係獨立且 未經修正的。 在此情形下，每個傳輸器......N的整體不確定性便 能以中位數的分量 <1 > < 2 > Μ^ζ = μ . Α (4)v, C Combining uncertainty The uncertainty of the overall model is a combination of model uncertainty and measurement noise. In order to obtain a simple model, it is assumed here that the uncertainty is independent and uncorrected. In this case, the overall uncertainty of each transmitter ... N can be expressed as a median component < 1 > < 2 > Μ ^ ζ = μ. Α (4)
V，< v，C 以及標準差 I<T> <2> , Ή(σν，+(σν， ⑺ 來描述。 定位階段（ISO) 在定位階段1 5 0中，複數個終端應可被同時地定位。 對於由定位所決定的服務而言，由於位置資訊係在伺服 端而不是在客戶端獲得，因此定位必須盡可能地在儲存了該 映照資訊的一中央電腦上執行。 對於定位演算來說其會產生下列需求。 -16- 200525167V, < v, C and standard deviation I < T > < 2 >, Ή (σν, + (σν, ⑺) are described. Positioning phase (ISO) In the positioning phase 1 50, a plurality of terminals should be available Simultaneous positioning. For services determined by positioning, since the location information is obtained on the server rather than on the client, positioning must be performed on a central computer that stores the mapping information as much as possible. For positioning calculations It will generate the following requirements. -16- 200525167
電腦容量必須盡可能地分佈於不同的定位應用之間，因 M 此需要具有任何時刻皆可調整精確性的一演算法。 · 這種演算法每次皆會提供一有意義的結果，計算上不成 熟的結果因此只會影響到位置估計的精確性而非其可利用 性，估計的精確性可藉由一較長的運作時間而增加。 基於場強度映照，在位置估計期間一非線性會產生，因 此必須使用一非線性濾除方法。 其會額外使用另一模型於使用者的移動以改善估計品 質並使得測量得以被結合。 φ 應用於定位之適當的濾除方法包括可找出一最佳解答 . 的漸進式背氏濾波器（引用文獻[7])或是PDSME濾波器（引用 文獻[5])。 雖然後者較爲次等，但其卻較爲簡單並且已在無線網路 中被用來測試定位問題。 對於定位一行動通訊裝置來說，首先會執行待定位置上 一次或多次的電場強度測量。 使用這些測量和前面所產生的場強度映照可執行一定 β 位估計。 爲了這個目的，舉例來說，亦可搭配場強度映照的物理 樣本而使用引用文獻[12]的模式匹配方法、引用文獻[5]的非 線性濾波方法或是導航方程式的解法。 使用位置向量基地台之所測量的場強度—五^%或 1 Ν 是天線1 - Ν、以及場強度Ε 1 ( X) - Ε Ν ( λ：)之定位所決定的映照或 模型；例如，藉由一零搜尋方法或是方陣方程式錯誤總和 -17- (6) 200525167 e ( 7 )的最小化可以解出導航方程式（6 )。 Εχ(χ) - E^eSS Ε2(Χ) - E^eSS ^ x ^mess EN(x) - En e = Σ (Ei(x) - E^ess)2 . (7) i = l 3.實驗證明 爲了證明上述程序，首先在一無線局部區域網路、亦即 所謂的DUKATH網路（引用文獻[8])中執行測量。 由於我們有興趣的是提供由定位所決定的服務’例如傳 輸備份資料至PDA，因此必須提供能夠針對大量的使用者同 時進行定位的可能性。 然而，其結果並不侷限於無線局部區域網路或是 DUKATH網路，其亦可被傳輸至其他的無線網路’例如藍芽 或D E C T標準的網路。 技術設計 藉由一自主型行動自動控制裝置（第4圖的400或1 10) 可執行映照建構。 引用文獻[9]的OmniBase(泛基地）平臺可用以達成這個 目的。 對於映照來說，該平臺具有一特殊的設計（如第4圖）， 其上設置了一台具有L u c e n t Ο r i η 〇 c 〇無線局部區域網路卡 之 Compaq 的 iPAQ Pocket PC401 〇 -18- 200525167 當自動控制裝置4 0 0通過該建築物時，i p A Q 4 0 1會測量 所有可用之基地台的場強度。 在一 IEEE8〇2.11(引用文獻[10])的網路中，基地台會持 續傳輸指標信號（beacon Signal)，行動終端可使用這些信號 以確認它們藉由獲得最佳連結的基地台，藉由查詢場強度可 對於這些指標信號展開評估。 基地台所隸屬的網路在這裡是無關的，用於測量的網路 卡在這裡的表現是被動的；也就是說，其本身並不發出任 何信號。 測量結果係經由一串列介面而被傳輸至自動控制裝置 4 00，而在其內與自動控制裝置所估計之參考位置一起進行 進一步的處理。 自動控制裝置定位 泛基地平臺 泛基地全方向自動控制平臺具有模組化的設計，其具有 四個類似的轉輪模組，每個模組皆具有一標準轉輪。 每個轉輪模組具有兩個驅動馬達，其一係用以定向，另 一則係用以驅動轉輪，由於總共的八個轉輪模組馬達被彼此 獨立地控制，因此可以自由地將該平臺以三種自由度移動 (位置及方向）。 該平臺係藉由一低階控制器協調驅動，該平臺永遠圍繞 所謂的瞬時旋轉中心（ICR， Instantaneous Center of R o t a t i ο η )進行瞬時旋轉移動。 爲了達成低振動行進，四個轉輪必須在IC R周圍的匹配 200525167 路徑上進行補償。 路徑設計係以一高階控制電腦執行，該控制電腦和該平 臺的低階控制器之間的通訊係經由一 CORBA介面產生。 其提供三種函數： -於全域座標中詢問自動控制裝置的里程估計位置 -於全域座標中設定位置 -設定速度 速度向量uk = 詳細說明了全域座標系中沿著 X軸和y軸的標稱速度以及自動控制裝置之方向的角向量。 該平臺便因此而被控制在全域座標中。 影像處理 自動控制裝置400本身係藉由使用空間內一致性之頂部 型態的一頂部攝影機402而定位其本身。 該頂部可由1 200mm2的白色頂部控制面板所構成，在面 板之間是灰色支架，頂部支架之間的交叉係以標記來標示， 彼此間具有1 2 0 0 m m的間隔。 爲了偵測標記，首先便須使用 Intel Open Source Computer Visin Library(引用文獻[11])的一角濾、波器（corner filter)。 四個角點所構成、由20至35個畫素的邊長所形成的正 方形會被偵測爲一標記，當邊長實際上大約相等（公差：2畫 素）及角度大約爲90度時，該正方形便可以包括角點的方式 表示。 第5圖係爲用以偵測標記的一攝影機影像5 00。 -20- 200525167 由該彳示記的貫際位置以及其在攝影機影像中的位置便 可獲得自動控制裝置的測量位置。 定位 由於自動控制裝置的位置係取決於影像分析中的全域 座標，因此測量方程式可以簡化如下： k + vk (8) 系統錯誤測量錯誤v k係假設爲未經修改且平均分佈於 具有協方差R的中位値零的附近。 在每一時間步驟中皆會執行一預測步驟，藉此由里程決 定自動控制裝置的預測位置if，錯誤協方差可簡化爲 P e C; =C (9) k k-\ 其中Q爲系統錯誤的協方差矩陣，Cp爲前一個、Ce爲 次一個的協方差矩陣。 每當一測量資料項目出現；亦即一標記被偵測到，則 須執行一測量步驟，但這並非出現於每一時間步驟中，這是 因爲簡化的測量方程式、所謂的Kalman增益Κκ可以下列簡 式表示 KK=c»y' (ίο) k k e P P X = X +Kk'{zk-x k k k k k 自動控制裝置的估計位置\ k (11) 和次一*個協方差如下 以及 (12)The computer capacity must be distributed between different positioning applications as much as possible, so M needs an algorithm that can adjust the accuracy at any time. · This algorithm will provide a meaningful result every time. The immature results of calculation will therefore only affect the accuracy of the position estimation rather than its availability. The accuracy of the estimation can be achieved by a longer operation. Increase with time. Based on the field intensity mapping, a non-linearity is generated during the position estimation, so a non-linear filtering method must be used. It will additionally use another model on the user's movement to improve the estimated quality and allow the measurements to be combined. Appropriate filtering methods for φ positioning include a progressive back filter (reference [7]) or a PDSME filter (reference [5]) that can find an optimal solution. Although the latter is inferior, it is simpler and has been used to test location problems in wireless networks. For locating a mobile communication device, one or more electric field strength measurements at a pending position are performed first. Using these measurements and the field intensity mappings previously generated, a certain β-bit estimation can be performed. For this purpose, for example, the pattern matching method cited in [12], the non-linear filtering method cited in [5], or the solution of the navigation equation can also be used with the physical samples mapped by field intensity. Measured field strength using a position vector base station-5 ^% or 1 Ν is the mapping or model determined by the positioning of the antenna 1-Ν and the field strength Ε 1 (X)-Ε Ν (λ :); for example, The navigation equation (6) can be solved by a zero search method or by minimizing the error sum of the square matrix equation -17- (6) 200525167 e (7). Εχ (χ)-E ^ eSS Ε2 (χ)-E ^ eSS ^ x ^ mess EN (x)-En e = Σ (Ei (x)-E ^ ess) 2. (7) i = l 3.Experiment Proof In order to prove the above procedure, measurements are first performed in a wireless local area network, also known as a DUKATH network (reference [8]). Since we are interested in providing positioning-determined services, such as transferring backup data to a PDA, we must provide the possibility of simultaneously positioning for a large number of users. However, the results are not limited to wireless local area networks or DUKATH networks, and they can also be transmitted to other wireless networks' such as Bluetooth or DE C T standard networks. Technical design Mapping construction can be performed by an autonomous mobile automatic control device (400 or 110 in Fig. 4). The OmniBase (pan-base) platform cited in [9] can be used to achieve this. For the mapping, the platform has a special design (as shown in Figure 4), which is equipped with a Compaq iPAQ Pocket PC401 with Lucent 〇 ri η 〇c 〇 wireless local area network card 〇-18- 200525167 When the automatic control device 400 passes the building, the ip AQ 401 measures the field strength of all available base stations. In an IEEE802.11 (reference [10]) network, base stations will continuously transmit beacon signals, and mobile terminals can use these signals to confirm that they are using the best-connected base station, Querying the field strength can evaluate these indicators. The network to which the base station belongs is irrelevant here, and the performance of the network card used for measurement is passive here; that is, it does not send any signal itself. The measurement results are transmitted to the automatic control device 400 via a serial interface, and further processed together with the reference position estimated by the automatic control device. Automatic control device positioning Pan-base platform Pan-base omnidirectional automatic control platform has a modular design, which has four similar runner modules, each module has a standard runner. Each runner module has two drive motors, one for orientation and the other for driving the runners. Since a total of eight runner module motors are controlled independently of each other, the The platform moves in three degrees of freedom (position and orientation). The platform is coordinately driven by a low-order controller, and the platform always performs instantaneous rotational movement around a so-called Instantaneous Center of Rot a t i τ η. To achieve low vibration travel, the four runners must be compensated on a matching 200525167 path around IC R. The path design is performed by a high-level control computer, and the communication between the control computer and the low-level controller of the platform is generated through a CORBA interface. It provides three functions:-Asking for the estimated mileage position of the automatic control device in the global coordinates-Setting the position in the global coordinates-Setting the speed and speed vector uk = specifies the nominal speed along the X and y axes in the global coordinate system And the angular vector of the direction of the automatic control device. The platform is thus controlled in global coordinates. The image processing automatic control device 400 itself locates itself by using a top camera 402 with a consistent top configuration in space. The top can be composed of a white top control panel of 1,200 mm2, with gray brackets between the panels, and the crosses between the top brackets are marked with marks, with a space of 1 200 mm between each other. In order to detect the mark, the corner filter and corner filter of the Intel Open Source Computer Visin Library (reference [11]) must be used first. A square formed by four corner points and formed by the side length of 20 to 35 pixels will be detected as a mark. When the side lengths are approximately equal (tolerance: 2 pixels) and the angle is approximately 90 degrees, The square can be represented by including corner points. FIG. 5 is a camera image 500 for detecting a mark. -20- 200525167 The measurement position of the automatic control device can be obtained from the interposition position of this indication and its position in the camera image. Positioning Since the position of the automatic control device depends on the global coordinates in the image analysis, the measurement equation can be simplified as follows: k + vk (8) System error The measurement error vk is assumed to be unmodified and evenly distributed over the covariance R Near the median zero. At each time step, a prediction step is performed, whereby the predicted position of the automatic control device is determined by the mileage, and the error covariance can be simplified to P e C; = C (9) k k- \ where Q is a system error The covariance matrix of C, where Cp is the covariance matrix of the previous one and Ce is the next. Each time a measurement data item appears; that is, a marker is detected, a measurement step must be performed, but this does not occur in every time step, because the simplified measurement equation, the so-called Kalman gain κ, can be Expressed in short form KK = c »y '(ίο) kke PPX = X + Kk' (zk-x kkkkk the estimated position of the automatic control device \ k (11) and the next one * covariance are as follows and (12)
e p Pe p P
C -Kk.C k K k k 自動控制裝置/的估計位置接著便可被設定成平臺上 k 200525167 的目前位置。 第6圖表示航行位移計算（dead reckoning)與攝影機領 航（camera navigation)之誤差的比較圖。 爲了這個目的，在一無線網路中一建築物的一迴廊會被 橫越10次，以對應於大約700m的總距離，在10次的橫越 之後，方向上的誤差大於4 5度，位置上的誤差則是數公尺’ 如果加入影像處理，平均位置估計誤差則約爲1 5 c m。 映照建構 對於映照建構來說，方程式3的6個參數係針對每一個 基地台進行估計，自動控制裝置的參考位置係用於xk，位於 位置之傳輸器f的測量場強度係用於ht(xk)。 使用最少次的平方法即可進行估計。 第7圖和第8圖表示兩個基地台的次要映照。 此外，實際的測量値以及自動控制裝置的移動路徑皆已 繪出，第7圖和第8圖係爲實際測量値的模型。 藉由和測量値有關的該測量模型系統的誤差便能夠估 計系統的整體不確定性，而部分的不確定性仍爲未知，藉由 諸如計算一點上的複數個測量的中位數及標準差便可以決 定測量雜訊，然而，由於此處的測量係沿著不同路徑執行， 因此一般來說，每個路徑點僅有一個測量。 由於在某些情形下，無線網路的傳輸基地台與自動控制 裝置係分別位於建築物的不同樓層，因此，藉由延伸位置向 量以形成一四維的位置向量在某些情形下可以增加模型的 正確性。 在此情形下，Z可用以補強高度，而0係用於χ-y平面 -22- 200525167 的定向。 其亦可將待估計之參數的數目從6N增加到1 5N，因此 · 需要有更多的測量値用於映照建構，爲了這個目的，測量動 作便可理想化地執行於其他樓層。 本案亦提出一種系統，其可自動檢視及映照無線網路的 局部信號特性。 該系統的一種應用係建立可用於定位無線網路中之使 用者的映照，所產生的映照係以分析型態描述所有可用之傳 輸基地台的場強度分量。 Φ 模型錯誤和測量雜訊係以隨機方式建立模型，分析式隨 . 機濾波方法因此能夠基於這些映照而被使用於定位。 在無線網路中使用泛基地全方向性自動控制裝置平臺 可執行本案之測量。 即使本案發明係藉由以上之較佳實施例來作說明，然而 對於熟習本項技術者來說，本案仍不限於這些實施例和使用 方法，尤有甚者，凡依本案所附申請專利範圍所做的均等變 化及修飾，皆爲本案專利範圍所涵蓋。 ® 【本案引用文獻如下】 [1] Peyrard, F., Soutou, C.? Mercier, J.J.: Mobile Stations Localization in a WLAN, in: Proceedings of the 25th Annul IEEE Conference on Local Computer Networks (LCNOO), Tampa, Florida (2000) 136-142 [2] Hassan-Ali, M·，Pahlavan, K.: A New Statistical Model for Site-Specific Indoor Radio Propagation Prediction Based on Geometric Optics -23- 200525167 and Geometric Probability. IEEE Transactions on Wireless Communications 1 (2002) 112-124 [3] Howard，A.，Siddigi，S·，Sukhatme, G.S.: An Experimental Study of Localization Using Wireless Ethernet. Appear in: Proceedings of the 4th International Conference on Field and Service Robotics, Japan (2003) [4] Bahl, P., Padmanabhan, V.N.: RADAR: An In-Building RF-based User Localization and Tracking System. In: Proceedings of IEEE INFOCOM 2000. Volume 2·，Tel Aviv，Israel (2000) 775-784 [5] Rauh，A.，Briechle，K.，Hanebeck，U.D·，Bamberger, J·，Hoffman, C·: Localization of DECT Mobile Phones Based on a New Nonlinear Filtering Technique. In: Proceedings of SPIE Bd. 5084，AeroSense Symposium， Orlando, Florida (2003) [6] Fleury，B.H., Leuthold, P.E.: Radiowave Propagation in Mobile Communications: An Overview of European Research. IEEE Communications Magazine 34 (1996) 70-81 [7] Hanebeck, U.D.，Briechle, K·，Rauh，A.，： Progressive Bayes: Anew Framework for Nonlinear State Estimation. In: Proceedings of SPIE Bd. 5099, AeroSense Symposium, Orlando, Florida (2003) [8] Hettler, A., Wigand, R.: DUKATH-Drahtlose Universitat [Wireless University] Karlsruhe (TH) (Status: July 2003) http://www.uni-karlsruhe.de/_DUKATH/.The estimated position of C -Kk.C k K k k automatic control device / can then be set to the current position of k 200525167 on the platform. Figure 6 shows a comparison of the error between dead reckoning and camera navigation. For this purpose, a corridor of a building in a wireless network will be traversed 10 times to correspond to a total distance of about 700m. After 10 traversals, the direction error is greater than 45 degrees. The error is several meters'. If image processing is added, the average position estimation error is about 15 cm. Mapping construction For mapping construction, the 6 parameters of Equation 3 are estimated for each base station, the reference position of the automatic control device is used for xk, and the measurement field strength of the transmitter f located at the position is used for ht (xk ). Estimates can be made using the least squares method. Figures 7 and 8 show the secondary mapping of the two base stations. In addition, the actual measurement frame and the movement path of the automatic control device have been drawn. Figures 7 and 8 are models of actual measurement frame. With the measurement system error related to the measurement 値, the overall uncertainty of the system can be estimated, while part of the uncertainty is still unknown. For example, by calculating the median and standard deviation of multiple measurements at one point You can decide to measure the noise, however, since the measurements here are performed along different paths, generally there is only one measurement per path point. In some cases, the transmission base station and automatic control device of the wireless network are located on different floors of the building. Therefore, by extending the position vector to form a four-dimensional position vector, the model can be added in some cases. Correctness. In this case, Z can be used to reinforce the height, and 0 is used for the orientation of the χ-y plane -22- 200525167. It can also increase the number of parameters to be estimated from 6N to 15N, so more measurements are needed (for mapping construction). For this purpose, the measurement can be ideally performed on other floors. This case also proposes a system that can automatically view and map local signal characteristics of wireless networks. One application of the system is to establish mappings that can be used to locate users in a wireless network. The resulting mappings describe the field strength components of all available transmission base stations in an analytical manner. Φ Model errors and measurement noise are modeled in a random manner. The analytical random filtering method can therefore be used for positioning based on these mappings. The use of a pan-base omnidirectional automatic control device platform in a wireless network can perform the measurement in this case. Even though the invention of this case is explained by the above preferred embodiments, for those skilled in the art, this case is still not limited to these embodiments and methods of use, especially, the scope of patents attached to this case All equal changes and modifications are covered by the patent scope of this case. ® [The references cited in this case are as follows] [1] Peyrard, F., Soutou, C.? Mercier, JJ: Mobile Stations Localization in a WLAN, in: Proceedings of the 25th Annul IEEE Conference on Local Computer Networks (LCNOO), Tampa, Florida (2000) 136-142 [2] Hassan-Ali, M ·, Pahlavan, K .: A New Statistical Model for Site-Specific Indoor Radio Propagation Prediction Based on Geometric Optics -23- 200525167 and Geometric Probability. IEEE on Wireless Communications 1 (2002) 112-124 [3] Howard, A., Siddigi, S., Sukhatme, GS: An Experimental Study of Localization Using Wireless Ethernet. Appear in: Proceedings of the 4th International Conference on Field and Service Robotics, Japan (2003) [4] Bahl, P., Padmanabhan, VN: RADAR: An In-Building RF-based User Localization and Tracking System. In: Proceedings of IEEE INFOCOM 2000. Volume 2. ·, Tel Aviv, Israel (2000) 775 -784 [5] Rauh, A., Briechle, K., Hanebeck, UD ·, Bamberger, J ·, Hoffman, C ·: Localization of DECT Mobile Phones Ba sed on a New Nonlinear Filtering Technique. In: Proceedings of SPIE Bd. 5084, AeroSense Symposium, Orlando, Florida (2003) [6] Fleury, BH, Leuthold, PE: Radiowave Propagation in Mobile Communications: An Overview of European Research. IEEE Communications Magazine 34 (1996) 70-81 [7] Hanebeck, UD, Briechle, K ·, Rauh, A.,: Progressive Bayes: Anew Framework for Nonlinear State Estimation. In: Proceedings of SPIE Bd. 5099, AeroSense Symposium, Orlando , Florida (2003) [8] Hettler, A., Wigand, R .: DUKATH-Drahtlose Universitat [Wireless University] Karlsruhe (TH) (Status: July 2003) http://www.uni-karlsruhe.de/_DUKATH/ .
[9] Hanebeck, U.D., Saldic, N., Freyberger, F., Schmidt, G.: Modulare Radsatzsysteme fur omnidirektionale mobile Roboter [Modular wheelset systems for omnidirectional mobile robots]. In: Robotik 2000 Conference 200525167 (VDI/VDE Gesellschaft Mess- und Automatisierungstechnik), VDI Berichte 1552, Berlin (2000) 39-44 [10]IEEE: Std 802.1 1-1997 Information Technology-Telecommunications And Information Exchange Between Systems-Local And Metropolitan Area Networks-Specific Requirements-Part 11: Wireless LAN Medium Access Control (MAC) And Physical Layer (PHY) Specifications. (1997) [lljlntel Corporation: Open Source Computer Vision Library. (Status: July 2003) http://www.intel.com/research/mrl/research/opencv/ [12]DE Patent application with filing reference 10345255.9 【圖式簡單說明】 第1圖表示本案藉由後續定位（subsequent localization) 產生場強度映照之程序的一實施例； 第 2 圖表示本案藉由同步定位（simultaneous localization)產生場強度映照之程序的一實施例； 第3圖表示模型錯誤和測量雜訊的圖形比較·； 第 4 圖表示本案全方向性自動控制裝置（〇mni directional robot)的一實施例； 第5圖表示使用影像處理所得的位置偵測； 第6圖表示以航行位移計算（dead reckoning)和攝影機 領航（c a m e 1· a n a v i g a t i ο η)之位置估計的圖形比較； 第7圖表示本案應用於無線網路一第一基地台之一第一 (&要）映照的一實施例；以及 第8圖表不本案應用於無線網路一第二基地台之一第二 (次要）映照的一實施例。 200525167 【元件符號說明】 1 00 映照階段 110' 400 自動控制裝置 111 測量 112 參考位置 115 映照建構 120 映照 12 1 次要映照 1 50 定位階段 160 使用者 165 定位 170 位置估計 401 iPAQ 402 頂部攝影機 500 攝影機影像 -26-[9] Hanebeck, UD, Saldic, N., Freyberger, F., Schmidt, G .: Modulare Radsatzsysteme fur omnidirektionale mobile Roboter [Modular wheelset systems for omnidirectional mobile robots]. In: Robotik 2000 Conference 200525167 (VDI / VDE Gesellschaft Mess -und Automatisierungstechnik), VDI Berichte 1552, Berlin (2000) 39-44 [10] IEEE: Std 802.1 1-1997 Information Technology-Telecommunications And Information Exchange Between Systems-Local And Metropolitan Area Networks-Specific Requirements-Part 11: Wireless LAN Medium Access Control (MAC) And Physical Layer (PHY) Specifications. (1997) [lljlntel Corporation: Open Source Computer Vision Library. (Status: July 2003) http://www.intel.com/research/mrl/research/opencv / [12] DE Patent application with filing reference 10345255.9 [Schematic description] Figure 1 shows an embodiment of the procedure for generating field intensity mapping by subsequent localization; Figure 2 shows the case by synchronous localization (Simultaneous localization) An example of the procedure according to the procedure; FIG. 3 shows a graphical comparison of model errors and measurement noise; FIG. 4 shows an embodiment of the omni-directional automatic control device (〇mni directional robot) in this case; FIG. 5 shows the use of images Processed position detection; Figure 6 shows a graphical comparison of the position estimates based on dead reckoning and camera 1 (anavigati ο η); Figure 7 shows the application of this case to wireless network An embodiment of the first (& required) mapping of one of the base stations; and an embodiment in which the eighth chart is applied to a second (secondary) mapping of a second base station of the wireless network. 200525167 [Description of component symbols] 1 00 mapping stage 110 '400 automatic control device 111 measurement 112 reference position 115 mapping construction 120 mapping 12 1 secondary mapping 1 50 positioning phase 160 user 165 positioning 170 position estimation 401 iPAQ 402 top camera 500 camera Image-26-