Patent Application
20050282537
This invention relates to improving wireless measurements and more particularly to systems and methods for increasing deployment of wireless measurement probes.
In all measurement problems, basic sampling criteria must be satisfied to understand the underlying physical phenomena of interest. It is widely accepted that a sample rate greater than twice that of the highest frequency of interest in the measured signal is required. This is known as the Nyquist criteria.
For measuring signals over wide geographical areas it is important to both sample the signal at many spaced-apart locations and to sample those signals at a sufficiently high rate so as to satisfy the Nyquist sampling rate criteria. For many systems, such as, by way of example, cellular RF signals from a cell site, probes mounted in (or on) vehicles are used. However, when probes are mobile, the spatial and temporal constraints interact with each other. A mobile probe often will have a sampling policy that is triggered by a) if moving fast, after a certain amount of distance has been traversed, trigger the next measurement, or b) if moving slow or not at all, after a certain amount of time has elapsed, trigger the next measurement.
Measurement problems typically involve the following steps: data acquisition, scaling and conversion, analysis, storage and presentation. Many traditional instruments are available that perform one or more (and typically all) of these functions within a single device. An operator interacts with these traditional instruments via a “front panel” or via a PC to configure the measurement conditions and view the measurement results.
Using wireless technology, the measurement probe can be located at a location remote from a user with measurement results steaming across the RF link to the user. This approach can be expanded to large-scale measurement problems with many probes communicating over RF and/or fixed-line links to a central data center.
Such remotely operated systems are highly accurate within the area they cover and are typically less costly than single-operation systems. However, the wireless remote probes are expensive and thus it is not economical to have a large number of such probes deployed. The result, then, is an insufficient number of probes being deployed to meet the sampling criteria discussed above.
Another cost factor is installation costs. Typically, the wireless probe would be permanently mounted in a vehicle (taxi, bus, company truck, etc.). This involves mounting RF, and possibly GPS, antennas on the roof of the vehicle, mounting the probe electronics in a secure location, such as in a trunk, then running coax from the antenna(s) to the probe, tapping into the vehicle's power supply (so that the probe knows if the vehicle ignition is turned on,) and sometimes tapping into the vehicle's transmission pulse circuit to assist with GPS dead-reckoning, etc.
Using existing probe designs, the system designer is stuck in a classic “catch-22.” This arises from the dilemma that probes that make powerful measurements are expensive and because they are expensive the designer can't afford too many of them. Since only a limited number of probes are affordable there are fewer of them than required, and thus each probe must be powerful in terms of measurement capabilities. Since each probe must be powerful, they are even more expensive, leading to a further reduction in the number of probes used in a system.
In one embodiment, a system and method is arranged so that less expensive probes are designed for simple installation. Since both probe cost and installation cost are reduced more probes can be deployed thereby improving wireless measurements across a system. In one embodiment, a self-contained probe is designed using internally generated power and vehicle speed information thereby eliminating the need for connecting into a vehicle's power. The low-cost design of each probe allows the electronics to be externally mounted in a quick and easy manner. The resulting large number of deployed probes then allows for different measurements being sent from different probes thereby further reducing the cost per probe.
The micro-wind turbine should be designed to minimize vibration and noise transmitted through the vehicle roof. With the power available from the wind being proportional to the collection area (i.e. the square of the rotor diameter in a conventional wind generator), an efficient design would be to maximize the frontal area of the turbine. As shown, a squirrel cage design is envisioned because it can be safely recessed into a plastic weather-proof housing. Input air slot 23 (
Because wind power increases with the cube of the wind velocity, wind energy is a significant power source on a rapidly moving vehicle (e.g. 80 km/hr or greater). For example, about four times more wind energy is available when traveling at 80 km/hr (50 mph) compared to 50 km/hr (30 mph). Of course at high speeds, measurement sampling rates are also high (short duration), so this extra power is put to good use.
To accommodate multiple sensors, measurement electronics 17 should be partitioned between the measurement front/end (i.e. STIM) and measurement backend (i.e. NCAP) as outlined in the EEEE 1451.2 standard. Measurement of the speed of rotation of wind-powered micro turbine generator 18 is used to estimate the vehicle speed (and thus location) during periods of no GPS signal. This internally generated speed generation capability is integrated into the GPS dead-reckoning mechanism and eliminates the need to run a cable to the vehicle's transmission pulse detection circuit.
Probe housing 11, in one embodiment, has air slot 21 (
Because of the relatively low cost of each probe and the low (almost non-existent) installation cost, there can be a large number of measurement probes deployed in a system so each probe only needs to satisfy only a small part of the overall measurement task. Accordingly, the overall measurement task is divided into small pieces with lots of probes working to acquire the necessary total task measurements. In addition, the probes are designed to interact with higher-level servers at a control location, not with local users. Therefore, local keyboards and displays are eliminated, further simplifying the design.
To illustrate how a swarm of inexpensive probes can work together, consider the measurement and diagnostics of hand-off problems in today's cellular networks. In cellular and 802.x networks, the network is typically partitioned into cells. As the wireless devices travel through the network they undergo a series of handoff operations between adjacent cells. In practice, numerous handoffs occur in the network during periods of high network usage. Having better measurements on the frequency and scope of handoff problems is an example of a high-value measurement for today's systems.
When attacking this problem with 100s or 1000s of inexpensive probes, each probe would measure and record only a few handoff measurement artifacts. When network bandwidth is available, (or when a mobile probe comes within range of a collection point) that probe transmits its stored measurements to a higher-level server. The higher-level source, such as control 42, can aggregate the data from a large number probes and analyze that data for signal problems.
One problem is how to configure each probe to make the measurement of interest. An approach is to use a statistical system such that each probe only measures a sub-set of the possible test measurements, but taken together over all of probes, all the desired test measurements are obtained. One system for implementing such an approach is to have every probe generate a random number (e.g. between 0 and 1) via, for example, random number generator (RNG) 103 (
Once the probe is simplified, its energy efficiency improves. The goal is to get the probe's energy requirements below the energy available from the integrated photovoltaic and/or micro-turbine wind generator. Rechargeable batteries allow the probe to operate during periods when generated power is not available.
While the embodiments discussed contain self-generated internal power, it is possible to obtain external power simply when such power is available on the roof of a vehicle. This would be true for police cars, taxis, maintenance trucks, etc. The design goal is to eliminate invasive attachment by eliminating the need to obtain power, or other signals, from the interior of the vehicle.
In one embodiment, one or more probes can function as a data recorder and the test results stored, for example, in a flash memory that can be removed for uploading to a database for processing. Alternatively, when the vehicle (or the test probe) is returned to a control location it can be physically linked (by wire or IR, etc.) for the upload of its information. In some situations, the “routine” test results could be communicated in this manner while the critical data could be transmitted to a control location.
The system and method discussed herein can, in addition to use for testing an RF signal, can be employed for any number of testing scenarios. For example, the device can have an opening for receiving air-borne substances, such as pollen, toxic material, radiation, etc., or any other type of input, such as sunlight, noise, vibration, images etc., for testing. As discussed, the results of these tests could be sent back immediately to a control point, or they could be stored in a memory in the device for subsequent uploading.
