CRAWDAD vanderbilt/interferometric

We collected localization traces from a radio interferometric tracking system, which is implemented on mote-class wireless sensor nodes.last modified : 2007-07-18release date : 2007-06-06date/time of measurement start : 2006-11-29date/time of measurement end : 2006-11-29collection environment : Location-awareness is an important requirement for many mobile wireless applications today. When GPS is not applicable because of the required precision and/or the resource constraints on the hardware platform, radio interferometric ranging may offer an alternative. In [kusy-intereferometric], we present a technique that enables the precise tracking of multiple wireless nodes simultaneously. To evaluate the performance of the technique, we use this dataset which was collected from a prototype implementation on mote-class wireless sensor nodes.network configuration : Our tracking system (called 'mTrack') is implemented using XSM motes from Crossbow, Inc. (a variant of the Berkeley Mica2 mote) as target and infrastructure nodes, as well as a PC laptop running the location computation and the tracking GUI. The motes are running TinyOS version 1.1.14 as the operating system. Our test environment is the empty Vanderbilt Football Stadium.data collection methodology : Sensors record the phase and frequency of the beat signal, while the transmitter pair iterates through a series of transmit frequencies. Q-ranges are calculated for every receiver pair and then converted to t-ranges, which are input to the analytical location solver. Once positions are known, q-speeds are converted to velocity vectors. Positions and velocities are displayed on a map and optionally used to control camera pitch/pan. (Please see [kusy-interferometric] for details about q-ranges, t-ranges, and q-speed.)Tracesetvanderbilt/interferometric/mtrackLocalization traceset collected from a radio interferometric tracking system.files: multipleMobile2hands.tar.gz, multipleMobile2hands.png, multipleMobile2handsa.tar.gz, 06-333-11-51-33.png, multipleMobile3.tar.gz, 06-333-11-50-16.pngdescription: Localization traceset collected from a radio interferometric tracking system which is implemented on Mica2 compatible XSM motes.measurement purpose: Location-aware Computing, Localizationmethodology: - Implementation The radio-interferometric ranging engine is implemented as an alternate radio driver. Since both the TinyOS radio stack and the ranging engine use the same radio hardware, the former is disabled during the measurement rounds. After a measurement round is completed, the results (phase and frequency measurements for each frequency channel) are routed to the PC using the Directed Flood Routing Framework (DFRF). DFRF is configured with the gradient convergecast policy to provide a fast and reliable data collection service in a multihop network. The computation of the interferometric q-ranges, their conversion to t-ranges, as well as the location computation are implemented in Java and run on a PC. (Please see [kusy-interferometric] for details about q-ranges and t-ranges.) The computation is implemented in a dataflow like manner, which would make it possible to distribute computational blocks to different computers, or to implement them in hardware. - Time Synchronization To assure that the phase measurements are carried out at the same time on all receivers, the receivers need to be time synchronized. Instead of choosing a synchronization service that maintains a global time in the whole network (such as FTSP), we opted for a multihop extension of the Estimated Time on Arrival (ETA) approach: the transmitter node at the common focus of the hyperbolae, called the master node, generates a beacon event tagged with the start time of the next measurement round by the node's local clock. ETA will propagate this beacon message to all nodes within a few radio hops, converting the timestamp to the local time of the recipients. ETA is able to achieve much better utilization of system resources than virtual global time services, because it does not synchronize the clock skews of different nodes. However, the beacon event and the start of the measurement round need to be close in time (300 ms), to achieve high synchronization accuracy despite the drifts of the unsynchronized local clocks at different nodes. Notice, that it is imperative that multihop time synchronization be used, because the maximum interferometric range exceeds the communication range of the motes. - Experimental Environment Our experimental hardware platform is a Mica2 compatible XSM mote, programmed using the nesC programming language and the TinyOS operating system. Our test environment is the empty Vanderbilt Football Stadium. - Measurement Details The mTrack system was able to get a position fix for each of the targets in approximately 4 seconds which includes 0.5 second coordination time, 1 second ranging time, 2 seconds multihop routing time, and 0.5 second localization time. During the coordination phase, infrastructure nodes were assigned the roles of transmitters and receivers and were time synchronized with each other and with the target nodes. The nodes measured the phase and frequency of the beat signal during the ranging phase and routed the measured values to the base station during the routing phase. Finally, a PC computer calculated the target locations and velocity vectors during the localization phase. We estimated the accuracy of mTrack's localization and velocity vector estimation with respect to the ground truth. Measuring the ground truth locations of multiple moving targets accurately, both spatially and temporally, however, was a difficult problem. We simplified this problem by predefining the tracks that the targets followed during the actual experiment. Each track consisted of a series of waypoints and the targets moved between two consecutive waypoints at an approximately constant speed on a straight line. During the actual experiment, we recorded the times at which each of the targets passed each of its predefined waypoints. This allowed us to compute the actual speed of the target for each segment of its predefined track. Moreover, since we also recorded the times when the ranging measurements were taken, we could determine the segment and interpolate the ground truth location of the target on that segment for any given ranging measurement. Therefore, for any given ranging measurement, we were able to reconstruct the ground truth location of the target as well as its velocity vector. - Experiment Scenario We deployed five anchor nodes at known surveyed locations, covering an area of approximately 27.4 - 27.4 m. We placed four anchors in the corners of this square and the fifth anchor close to the center. The actual setup can be seen in [Figure: the experimental setup for multipleMobile2hands(a)] and [Figure: the experimental setup for multipleMobile3], which shows the anchor node locations (as black dots), the track that a mobile node follows, and the calculated locations and velocity vectors. This dataset contains the data from the following three experiments: 1. multipleMobile2hands Simultaneous tracking of multiple nodes: a person holds two motes in two hands, approximately 1.5 m apart and walks on the rectangular track, second person holds a single mote and walks on the triangular track. The setup is shown in [Figure: the experimental setup for multipleMobile2hands]. 2. multipleMobile2handsa Same as multipleMobile2hands, but at a different time. The setup is shown in [Figure: the experimental setup for multipleMobile2handsa]. 3. multipleMobile3 Simultaneous tracking of multiple nodes: three different nodes are moving along 3 different tracks, starting at points A,B, and C. The setup is shown in [Figure: the experimental setup for multipleMobile3].vanderbilt/interferometric/mtrack TracesmultipleMobile2hands: Localization trace collected from a radio interferometric tracking system for tracking three mobile nodes.configuration: Simultaneous tracking of multiple nodes: a person holds two motes in two hands, approximately 1.5 m apart and walks on the rectangular track, second person holds a single mote and walks on the triangular track. The setup is shown in [Figure: the experimental setup for multipleMobile2hands]. Interferometric measurement involves 3 nodes: master, assistant, and receiver. Master and assistant transmit at the same time to create the interference measurement and the receiver measures frequency and phase of this interference signal. This is repeated at multiple channels. In the paper [kusy-interferometric], we used the following 22 channels: chan-ID 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 chan-freq(Mhz) 401.1409755 421.1528585 441.69137 417.9930875 438.531599 458.543482 402.1942325 422.2061155 442.744627 416.9398305 437.478342 457.490225 403.2474895 423.2593725 443.797884 415.8865735 436.425085 456.436968 404.3007465 424.3126295 444.851141 414.8333165 To measure the interferometric range (q-range), we need to take calculate phase offsets of two receivers at all 22 channels.format:.freq file:each line of this file stores the frequencies measured in one interference measurement, at one receiver node.In the experiments in the paper [kusy-interferometric], only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID freq@CH1 freq@CH2 freq@CH3 ...measurement-time is in hh:mm:ss.sssfreq@CHx means frequency in Hz measured at the radio channel with chan-ID 'x'.phase file:each line of this file stores the phases measured in one interference measurement, at one receiver node.In the experiments in [kusy-interferometric] experiments, only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID phase@CH1 phase@CH2 phase@CH3 ...measurement-time is in hh:mm:ss.sssphase@CHx means phase in radian measured at the radio channel with chan-ID 'x'.pos file:each line of this file stores information about a sensor node.parsing of each line: 'Sensor' moteID x y z is-anchor dontcare dontcare dontcaremoteID is ID of the motex is x-coordinate of the mote in metersy is y-coordinate of the mote in metersz is z-coordinate of the mote in metersis-anchor is one if the mote was anchor, zero if it was a tracked node.track file:this file stores the ground truth track for each of the tracked nodes. it consists of multiple sections, each of themstarting with #nodeID and continuing with the track data points: measurement-time x y zeach line should be interpreted as: at the measurement-time nodeID was located at (x,y,z).xls and .png files:these files show the output of our tracking algorithm (which takes frequencies, phases, and locations of anchor nodesand calculates interferometric ranges (q-ranges), q-speeds, t-ranges, and finally locations and velocities of thetracked nodes). excel file contains the data, png file shows the resulting figure published in the paper [kusy-interferometric]..set file:this file stores various settings, most notably IDs of the channels at which we generated the interference signal.channelID can be converted to MHz the following way: 430.105543 MHz + channelID*0.5266285 MHze.g. channelID1=-55 corresponds to 401.14 MHzmultipleMobile2handsa: Localization trace collected from a radio interferometric tracking system for tracking three mobile nodes.configuration: Simultaneous tracking of multiple nodes: a person holds two motes in two hands, approximately 1.5 m apart and walks on the rectangular track, second person holds a single mote and walks on the triangular track. The experimetal setup is the same as multipleMobile2hands, but was conducted at a different time. The setup is shown in [Figure: the experimental setup for multipleMobile2handsa]. Interferometric measurement involves 3 nodes: master, assistant, and receiver. Master and assistant transmit at the same time to create the interference measurement and the receiver measures frequency and phase of this interference signal. This is repeated at multiple channels. In the paper [kusy-interferometric], we used the following 22 channels: chan-ID 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 chan-freq(Mhz) 401.1409755 421.1528585 441.69137 417.9930875 438.531599 458.543482 402.1942325 422.2061155 442.744627 416.9398305 437.478342 457.490225 403.2474895 423.2593725 443.797884 415.8865735 436.425085 456.436968 404.3007465 424.3126295 444.851141 414.8333165 To measure the interferometric range (q-range), we need to take calculate phase offsets of two receivers at all 22 channels.format:.freq file:each line of this file stores the frequencies measured in one interference measurement, at one receiver node.In the experiments in the paper [kusy-interferometric], only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID freq@CH1 freq@CH2 freq@CH3 ...measurement-time is in hh:mm:ss.sssfreq@CHx means frequency in Hz measured at the radio channel with chan-ID 'x'.phase file:each line of this file stores the phases measured in one interference measurement, at one receiver node.In the experiments in the paper [kusy-interferometric], only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID phase@CH1 phase@CH2 phase@CH3 ...measurement-time is in hh:mm:ss.sssphase@CHx means phase in radian measured at the radio channel with chan-ID 'x'.pos file:each line of this file stores information about a sensor node.parsing of each line: 'Sensor' moteID x y z is-anchor dontcare dontcare dontcaremoteID is ID of the motex is x-coordinate of the mote in metersy is y-coordinate of the mote in metersz is z-coordinate of the mote in metersis-anchor is one if the mote was anchor, zero if it was a tracked node.track file:this file stores the ground truth track for each of the tracked nodes. it consists of multiple sections, each of themstarting with #nodeID and continuing with the track data points: measurement-time x y zeach line should be interpreted as: at the measurement-time nodeID was located at (x,y,z).xls and .png files:these files show the output of our tracking algorithm (which takes frequencies, phases, and locations of anchor nodesand calculates interferometric ranges (q-ranges), q-speeds, t-ranges, and finally locations and velocities of thetracked nodes). excel file contains the data, png file shows the resulting figure published in the paper [kusy-interferometric]..set file:this file stores various settings, most notably IDs of the channels at which we generated the interference signal.channelID can be converted to MHz the following way: 430.105543 MHz + channelID*0.5266285 MHze.g. channelID1=-55 corresponds to 401.14 MHzmultipleMobile3: Localization trace collected from a radio interferometric tracking system for tracking three mobile nodes.configuration: Simultaneous tracking of multiple nodes: three different nodes are moving along 3 different tracks, starting at points A,B, and C. The setup is shown in [Figure: the experimental setup for multipleMobile3]. Interferometric measurement involves 3 nodes: master, assistant, and receiver. Master and assistant transmit at the same time to create the interference measurement and the receiver measures frequency and phase of this interference signal. This is repeated at multiple channels. In the the [kusy-interferometric], we used the following 22 channels: chan-ID 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 chan-freq(Mhz) 401.1409755 421.1528585 441.69137 417.9930875 438.531599 458.543482 402.1942325 422.2061155 442.744627 416.9398305 437.478342 457.490225 403.2474895 423.2593725 443.797884 415.8865735 436.425085 456.436968 404.3007465 424.3126295 444.851141 414.8333165 To measure the interferometric range (q-range), we need to take calculate phase offsets of two receivers at all 22 channels.format:.freq file:each line of this file stores the frequencies measured in one interference measurement, at one receiver node.In the experiments in the paper [kusy-interferometric], only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID freq@CH1 freq@CH2 freq@CH3 ...measurement-time is in hh:mm:ss.sssfreq@CHx means frequency in Hz measured at the radio channel with chan-ID 'x'.phase file:each line of this file stores the phases measured in one interference measurement, at one receiver node.In the experiments in the paper [kusy-interferometric], only first 22 channels were measured, the values of the other channels are marked as NaN.parsing of each line: measurement-time seqNum masterID assistantID receiverID phase@CH1 phase@CH2 phase@CH3 ...measurement-time is in hh:mm:ss.sssphase@CHx means phase in radian measured at the radio channel with chan-ID 'x'.pos file:each line of this file stores information about a sensor node.parsing of each line: 'Sensor' moteID x y z is-anchor dontcare dontcare dontcaremoteID is ID of the motex is x-coordinate of the mote in metersy is y-coordinate of the mote in metersz is z-coordinate of the mote in metersis-anchor is one if the mote was anchor, zero if it was a tracked node.track file:this file stores the ground truth track for each of the tracked nodes. it consists of multiple sections, each of themstarting with #nodeID and continuing with the track data points: measurement-time x y zeach line should be interpreted as: at the measurement-time nodeID was located at (x,y,z).xls and .png files:these files show the output of our tracking algorithm (which takes frequencies, phases, and locations of anchor nodesand calculates interferometric ranges (q-ranges), q-speeds, t-ranges, and finally locations and velocities of thetracked nodes). excel file contains the data, png file shows the resulting figure published in the paper [kusy-interferometric]..set file:this file stores various settings, most notably IDs of the channels at which we generated the interference signal.channelID can be converted to MHz the following way: 430.105543 MHz + channelID*0.5266285 MHze.g. channelID1=-55 corresponds to 401.14 MHz本数据集采集自部署于mote级无线传感器节点(mote-class wireless sensor nodes)的无线电干涉跟踪系统(radio interferometric tracking system)的定位轨迹。 ### 元数据 最后修改时间：2007-07-18；发布日期：2007-06-06；测量开始时间：2006-11-29；测量结束时间：2006-11-29。 ### 采集背景 位置感知是当前诸多移动无线应用的重要需求。当全球定位系统(GPS, Global Positioning System)因精度要求或硬件平台资源限制无法适用时，无线电干涉测距(radio interferometric ranging)可作为替代方案。在文献[kusy-interferometric]中，我们提出了一种可同时对多个无线节点进行精准跟踪的技术。为评估该技术的性能，我们使用了本数据集，其采集自mote级无线传感器节点上的原型实现。 ### 网络配置 我们的跟踪系统名为"mTrack"，采用美国Crossbow公司(Crossbow, Inc.)的XSM mote作为目标节点与基础设施节点，该节点是伯克利Mica2 mote(Berkeley Mica2 mote)的变体；同时搭载运行位置计算与跟踪图形用户界面(GUI, Graphical User Interface)的笔记本电脑。节点运行TinyOS 1.1.14版本操作系统(TinyOS operating system)。实验环境为空旷的范德堡大学足球场(Vanderbilt Football Stadium)。 ### 数据采集方法论 传感器记录拍频信号(beat signal)的相位与频率，同时发射机遍历一系列发射频率。针对每一对接收机计算q-range(q-range)，并将其转换为t-range(t-range)，输入解析定位求解器。待获取位置后，将q-speed(q-speed)转换为速度矢量。位置与速度信息将显示在地图上，并可选择性用于控制摄像机的俯仰与摇摄。（关于q-range、t-range与q-speed的详细说明，请参见文献[kusy-interferometric]） ### 数据集总览 数据集标识：Tracesetvanderbilt/interferometric/mtrack 数据集描述：从部署于兼容Mica2的XSM mote的无线电干涉跟踪系统采集的定位轨迹集。 文件列表：multipleMobile2hands.tar.gz、multipleMobile2hands.png、multipleMobile2handsa.tar.gz、06-333-11-51-33.png、multipleMobile3.tar.gz、06-333-11-50-16.png 测量用途：位置感知计算、定位 ### 完整方法论 1. **系统实现**：无线电干涉测距引擎作为备用无线电驱动实现。由于TinyOS无线电协议栈与测距引擎共用同一无线电硬件，因此在测量回合期间会禁用前者。测量回合完成后，各频率信道的相位与频率测量结果将通过定向泛洪路由框架(Directed Flood Routing Framework, DFRF)路由至PC。DFRF配置为梯度汇聚播送策略(gradient convergecast policy)，可在多跳网络(multihop network)中提供快速可靠的数据采集服务。干涉测距q-range的计算、其向t-range的转换以及定位计算均通过Java实现并运行于PC。该计算采用类数据流的方式实现，可将计算模块分发至多台计算机，或通过硬件实现。（详细说明请参见文献[kusy-interferometric]） 2. **时间同步**：为确保所有接收机在同一时刻执行相位测量，需对接收机进行时间同步。我们未选择维持全网全局时间的同步服务（如FTSP），而是采用了基于到达估计时间(Estimated Time on Arrival, ETA)的多跳扩展方案：位于双曲线公共焦点的发射机节点（称为主节点）会生成一条信标事件，该事件携带由节点本地时钟标记的下一测量回合的开始时间。ETA将该信标消息传播至数跳范围内的所有节点，并将时间戳转换为接收方的本地时间。相较于虚拟全局时间服务，ETA能够更高效地利用系统资源，因为其无需同步不同节点的时钟偏移(clock skews)。但信标事件与测量回合的开始时间需保持在300ms以内，以在不同节点的未同步本地时钟存在漂移的情况下仍实现较高的同步精度。需注意，必须使用多跳时间同步，因为最大干涉测距范围超出了mote的通信范围。 3. **实验环境**：实验硬件平台为兼容Mica2的XSM mote，采用nesC编程语言(nesC programming language)与TinyOS操作系统进行编程。实验环境为空旷的范德堡大学足球场。 4. **测量细节**：mTrack系统可在约4秒内完成每个目标的位置解算，其中包含0.5秒的协调时间、1秒的测距时间、2秒的多跳路由时间以及0.5秒的定位时间。在协调阶段，基础设施节点会被分配为发射机与接收机，并彼此之间以及与目标节点进行时间同步。在测距阶段，节点测量拍频信号的相位与频率，并在路由阶段将测量值路由至基站。最后，PC在定位阶段计算目标位置与速度矢量。我们基于地面真值(ground truth)评估了mTrack的定位与速度矢量估计精度。然而，精准测量多个移动目标的时空地面真值位置是一项难题。我们通过预定义目标在实际实验中遵循的运动轨迹简化了该问题：每条轨迹由一系列航路点(waypoint)组成，目标以近似恒定的速度在直线上的两个连续航路点之间移动。在实际实验中，我们记录了每个目标经过其预定义航路点的时间，据此可计算出目标在预定义轨迹各段的实际速度。此外，由于我们还记录了测距测量的时间，因此可确定目标所处的轨迹段，并插值得到任意给定测距测量时刻的目标地面真值位置。因此，对于任意给定的测距测量，我们均可重构目标的地面真值位置及其速度矢量。 5. **实验场景**：我们在勘测确定的位置部署了5个锚节点(anchor node)，覆盖约27.4×27.4米的区域。其中4个锚节点部署于该正方形区域的四个角落，第5个锚节点部署于靠近中心的位置。实际部署设置可参见[图：multipleMobile2hands(a)的实验部署]与[图：multipleMobile3的实验部署]，其中展示了锚节点位置（黑色圆点）、移动节点的运动轨迹以及计算得到的位置与速度矢量。本数据集包含以下三个实验的数据： 1. **multipleMobile2hands**：多节点同时跟踪：一名受试者双手各持一个间距约1.5米的mote，沿矩形轨迹行走；第二名受试者手持一个mote，沿三角形轨迹行走。部署设置参见[图：multipleMobile2hands的实验部署]。 2. **multipleMobile2handsa**：与multipleMobile2hands实验设置相同，但在不同时间进行。部署设置参见[图：multipleMobile2handsa的实验部署]。 3. **multipleMobile3**：多节点同时跟踪：三个不同的节点沿三条不同的轨迹移动，分别从A、B、C点出发。部署设置参见[图：multipleMobile3的实验部署]。 ### 子数据集详细说明 #### vanderbilt/interferometric/mtrack Traces ##### multipleMobile2hands：用于跟踪三个移动节点的无线电干涉跟踪系统定位轨迹 配置：多节点同时跟踪：一名受试者双手各持一个间距约1.5米的mote，沿矩形轨迹行走；第二名受试者手持一个mote，沿三角形轨迹行走。部署设置参见[图：multipleMobile2hands的实验部署]。干涉测量涉及3个节点：主节点、辅助节点与接收机。主节点与辅助节点同时发射信号以产生干涉测量信号，接收机测量该干涉信号的频率与相位。该过程在多个信道上重复进行。在文献[kusy-interferometric]中，我们使用了以下22个信道： | chan-ID | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | | chan-freq(Mhz) | 401.1409755 | 421.1528585 | 441.69137 | 417.9930875 | 438.531599 | 458.543482 | 402.1942325 | 422.2061155 | 442.744627 | 416.9398305 | 437.478342 | 457.490225 | 403.2474895 | 423.2593725 | 443.797884 | 415.8865735 | 436.425085 | 456.436968 | 404.3007465 | 424.3126295 | 444.851141 | 414.8333165 | 为测量干涉测距(q-range)，需计算两个接收机在全部22个信道上的相位偏移。 **文件格式说明**： 1. **.freq文件**：每行存储一次干涉测量中单个接收机节点测得的频率值。在文献[kusy-interferometric]的实验中，仅测量了前22个信道，其余信道的值标记为NaN。每行解析格式：measurement-time seqNum masterID assistantID receiverID freq@CH1 freq@CH2 freq@CH3 ... 其中measurement-time格式为hh:mm:ss.sss；freq@CHx表示在信道ID为x的无线信道上测得的频率（单位：Hz）。 2. **.phase文件**：每行存储一次干涉测量中单个接收机节点测得的相位值。在文献[kusy-interferometric]的实验中，仅测量了前22个信道，其余信道的值标记为NaN。每行解析格式：measurement-time seqNum masterID assistantID receiverID phase@CH1 phase@CH2 phase@CH3 ... 其中measurement-time格式为hh:mm:ss.sss；phase@CHx表示在信道ID为x的无线信道上测得的相位（单位：弧度）。 3. **.pos文件**：每行存储一个传感器节点的信息。每行解析格式：'Sensor' moteID x y z is-anchor dontcare dontcare dontcare 其中moteID为mote的ID；x为mote的x坐标（单位：米）；y为mote的y坐标（单位：米）；z为mote的z坐标（单位：米）；is-anchor为1表示该节点为锚节点，为0表示该节点为被跟踪节点；dontcare为无关字段。 4. **.track文件**：存储每个被跟踪节点的地面真值轨迹。该文件包含多个区段，每个区段以#nodeID开头，后续为轨迹数据点，每行格式为：measurement-time x y z 含义为：在measurement-time时刻，节点ID为nodeID的节点位于(x,y,z)位置。 5. **.xls与.png文件**：展示我们的跟踪算法的输出结果（该算法以频率、相位与锚节点位置为输入，计算得到干涉测距(q-range)、q-speed、t-range，最终得到被跟踪节点的位置与速度矢量）。Excel文件包含数据，PNG文件展示了发表于文献[kusy-interferometric]中的结果图。 6. **.set文件**：存储各类设置，其中最关键的是生成干涉信号所用的信道ID。信道ID可通过以下公式转换为MHz：430.105543 MHz + channelID*0.5266285 MHz。例如，channelID=-55对应401.14 MHz。 ##### multipleMobile2handsa：用于跟踪三个移动节点的无线电干涉跟踪系统定位轨迹 配置：多节点同时跟踪：一名受试者双手各持一个间距约1.5米的mote，沿矩形轨迹行走；第二名受试者手持一个mote，沿三角形轨迹行走。实验设置与multipleMobile2hands相同，但在不同时间进行。部署设置参见[图：multipleMobile2handsa的实验部署]。干涉测量涉及3个节点：主节点、辅助节点与接收机。主节点与辅助节点同时发射信号以产生干涉测量信号，接收机测量该干涉信号的频率与相位。该过程在多个信道上重复进行。在文献[kusy-interferometric]中，我们使用了与multipleMobile2hands完全一致的22个信道。为测量干涉测距(q-range)，需计算两个接收机在全部22个信道上的相位偏移。 文件格式与multipleMobile2hands完全一致。 ##### multipleMobile3：用于跟踪三个移动节点的无线电干涉跟踪系统定位轨迹 配置：多节点同时跟踪：三个不同的节点沿三条不同的轨迹移动，分别从A、B、C点出发。部署设置参见[图：multipleMobile3的实验部署]。干涉测量涉及3个节点：主节点、辅助节点与接收机。主节点与辅助节点同时发射信号以产生干涉测量信号，接收机测量该干涉信号的频率与相位。该过程在多个信道上重复进行。在文献[kusy-interferometric]中，我们使用了与multipleMobile2hands完全一致的22个信道。为测量干涉测距(q-range)，需计算两个接收机在全部22个信道上的相位偏移。 文件格式与multipleMobile2hands完全一致。