Wireless rainfall sensor

Abstract

An apparatus and a method of measuring and / or monitoring rainfall. The apparatus for measuring rainfall intensity in a local geographic area includes an initial learning system and subsequent reporting means for reporting on a current correlation between a series of cellular signals and expected rainfall. The initial learning system includes a wireless receiver for receiving a series of cellular signals emitted from multiple base station signals in the neighbourhood of the wireless receiver; a monitoring unit for monitoring the rainfall in the region of the wireless receiver; and a machine learning unit for processing the correlation between the rainfall in a region and the corresponding attenuation of the cellular signals of multiple base stations for an extended period of time to produce a subsequent prediction model. The method of monitoring the rainfall over an extended period of time includes the steps of: (a) initially monitoring a plurality of cellular base station emissions to identify each base station location and signal characteristics; (b) extracting frequency and signal characteristics from the base station emissions, in conjunction with rainfall reports for the area around the monitoring system, utilising machine learning to produce an extended correlation model of the change in signal characteristics with respect to rainfall; and (c) subsequently utilising the extended correlation model to output a rainfall prediction based on the received frequency and signal characteristics.

Description

Wireless Rainfall SensorFIELD OF THE INVENTION

[0001] The present invention relates to rainfall sensors and in particular to those sensors which rely on electromagnetic signal attenuation to determine a rainfall amount.BACKGROUND OF THE INVENTION

[0002] Any discussion of the background art throughout the specification should in no way be considered as an admission that such art is widely known or forms part of common general knowledge in the field.

[0003] A varity of rainfall sensors are known. The traditional tipping bucket rain gauge relies on a bucket that tips and counts each time it fills with a fixed amount of rainfall. Optical rain sensors employ infrared or laser beams to detect raindrops and measure their size and frequency. Capacitive rain sensors use changes in capacitance to detect rainwater presence. Acoustic rain sensors rely on sound waves to measure raindrop impact, while ultrasonic sensors gauge rainfall by measuring the time it takes for sound waves to bounce off raindrops. All these rainfall sensors have certain disadvantages such as high cost in the units, installation and maintenance. As a result, these sensors are typically not deployed in large quantities, thus limiting the rainfall forecast accuracy.

[0004] Various attempts to utilise cellular base station signals for rainfall measurement have been made, including those discussed in the foillowing survey articles:

[0005] Lian, B.;Wei, Z.; Sun, X.; Li, Z.; Zhao, J. A Review on Rainfall Measurement Based on Commercial Microwave Links in Wireless Cellular Networks. Sensors 2022, 22, 4395.

[0006] Sakkas, A.; Christofilakis, V.; Lolis, C.J.; Chronopoulos, S.K.; Tatsis, G. Harnessing the Radio Frequency Power Level of Cellular Terminals for Weather Parameter Sensing. Electronics 2024, 13, 840.

[0007] S. -H. Fang and Y. -H. S. Yang, "The Impact of Weather Condition on Radio-Based Distance Estimation: A Case Study in GSM Networks With Mobile Measurements," in IEEE Transactions on Vehicular Technology, vol. 65, no. 8, pp. 6444-6453, Aug. 2016

[0008] However, generally the results have been poor and there is a general need for a higher quality measurement system.

[0009] Unless the context clearly requires otherwise, throughout the description and the claims, the words, “comprise”, “comprising”, and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to”.

[0010] References

[0011] [1] S. -H. Fang and Y. -H. S. Yang, "The Impact of Weather Condition on Radio-BasedDistance Estimation: A Case Study in GSM Networks with Mobile Measurements," in IEEE Transactions on Vehicular Technology, vol. 65, no. 8, pp. 6444-6453, Aug. 2016, doi: 10.1109 / TVT.2015.2479591.[2] https: / / www.3gpp.org /

[0012] [3] Ringner, M. What is principal component analysis? Nature Biotechnology 26, 303-304 (2008). https: / / doi.org / 10.1038 / nbt0308-303

[0013] [4] "Data Fusion: Concepts and Algorithms" by David L. Hall and James Llinas

[0014] [5] "Deep Learning" by Ian Goodfellow, Yoshua Bengio, and Aaron CourvilleSUMMARY OF THE INVENTION

[0015] It is an obj ect of the invention, in its preferred form to provide an improved form of rainfall sensor.

[0016] In some embodiments, there is provided an apparatus using cellular signals to measure the rainfall of a local area, including: a wireless receiver for receiving cellular signals from multiple base stations across multiple frequency bands; a signal processing unit that processes the above signals to obtain rainfall information; and a wireless transmitter that transmits the rainfall sensing results to an internet-based data management platform.

[0017] In some embodiments, the cellular signals used can be from either singular or multiple neighbouring base stations. These base stations can be 4G, 5G and 6G. The signals can include synchronization signals, broadcast pilots, cell-specific reference signals, and channel stateinformation signals. The cellular signals can be used from either singular or multiple frequency bands.

[0018] The embodiments also provide for a new method for data fusion and signal processing to obtain rainfall information.

[0019] The wireless transmitter used to report rainfall sensing results can utilise WiFi, cellular, Bluetooth, LoRaWAN, satellite etc.

[0020] The apparatus of the embodiments can also be used for detecting snowfall, hail and other weather factors having similar impact on microwave signals. In that case, with the overall sensing scheme remained the same, the machine learning models can be adapted to take into account the difference between the impacts of rainfall and snowfall on the propagation of wireless signals.

[0021] In some embodiments, the signal processing unit can be replaced with a central control unit which controls the reception of the wireless cellular signals and transmits the baseband signals to a cloud data management and computing platform using the wireless transmitter in Fig. 1. In this case, the signal processing methods and machine learning models in this embodiment are still applicable, except they are performed in an internet-based cloud platform.

[0022] In accordance with one aspect of the present invention there is provided an apparatus for measuring rainfall intensity in a local geographic area, the apparatus including: an initial learning system including: a wireless reciever for receiving a series of cellular signals emitted from multiple base station signals in the neighborhood of the wireless receiver; a monitoring unit for monitoring the rainfall in the reigon of the wireless receiver; a machine learning unit for processing the correlation between the rainfall in a region and the corresponding attenuation of the cellular signals of multiple base stations for an extended period of time; to produce a subsequent prediction model; and subsequent reporting means for reporting on a current correlation between the series of cellular signals and expected rainfall.

[0023] In accordance with one aspect of the present invention there is provided a method of monitoring the rainfall over an extended period of time, the mthod including the steps of: (a) initially monitoring a plurality of cellular base station emissions to identify each base station location and signal characteristics; (b) extracting frequency and signal characteristics from the base station emissions, in conjunction with rainfall reports for the area around the monitoring system, utilising machine learning to produce an extended correlation model of the change in signal characteristicswith respect to rainfall; and (c) subsequently utilising the extended correlation model to output a rainfall prediction based on the received frequency and signal characteristics.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

[0025] Fig. 1 illustrates a chart of the normalized received signal level (RSL) variance values for the three rainfall levels, over an observation time window.

[0026] Fig. 2 illustrates schematically the architecture of the wireless rainfall sensor

[0027] Fig. 3 illustrates a scenario diagram of a rainfall sensor based on in-field experiments inSydney, Australia

[0028] Fig. 4 illustrates matched fdtering results (in amplitude) of the synchronisation signals for different base stations with different cell identification numbers.

[0029] Fig. 5 illustrates exemplary key signal metrics of 8 cells after performing the processing sub-steps in Step 2, where all sub-figures share the same x-labels and y-labels. The ground-truth rainfall intensity, as obtained from [https: / / www.ecowitt.net / home / map], is provided against the right y-axis in each sub-figure as a reference to highlight the impact of rainfall on the extracted and processed signal metrics.

[0030] Fig. 6 shows the fused SINRs over all cells.

[0031] Fig. 7 illustrates the typical signal features for training machine learning models, which are extracted based on the fused signals as illustrated in Fig. 6.

[0032] Fig. 8 illustrates the rainfall sensing results employing a GPR model.

[0033] Fig. 9 illustrates a flowchart of the steps of one embodiment.DETAILED DESCRIPTION

[0034] The preferred embodiments disclose a rainfall sensor including a wireless rainfall sensor.

[0035] In comparison to other existing sensors, the preferred embodiments have a number of important features, for example: Non-Intrusive Measurement at low cost: The wireless rainfall sensor of the preferred embodiments don't require physical contact with the rain, or any mechanical parts like tipping buckets. This makes them more reliable and low cost. Continuous Monitoring: Wireless rainfall sensors of the preferred embodiments can provide continuous rainfall data without interruption. Robustness: Wireless rainfall sensors of the preferred embodiments are not affected by leaves and debris which tend to make other sensors malfunctioning. They are also more robust against environmental factors like wind, temperature, and humidity compared to some other sensor types. Accuracy: Since wireless rainfall sensors of the preferred embodiments monitor the rainfall in an area but not for a spot, the obtained information is much more accurate and useful.

[0036] The preferred embodiments thus disclose an innovative system architecture, method for information fusion, and signal processing methods for sensing rainfall, specifically measuring rain intensity and providing rainfall forecast. The embodiments can also be used for sensing snowfall. With the architecture and methods disclosed, a sensor can be made to sense both rainfall and snowfall.

[0037] The rainfall sensor utilizes the wireless signals over the access links of cellular communication systems. Such signals, which are transmitted by all existing cellular base stations, are received by a wireless receiver. The characteristics of these signals change according to the surrounding environment. In particular, if there is rainfall around the rainfall sensor, the received signal at the sensor will reflect the changes in the signals from neighbouring cellular base stations.

[0038] Fig. 1 illustrates an example of the impact of rain levels on the received signal level of a cellular signal, taken from the Yang et al citation.

[0039] By fusing and processing multiple received signals at the wireless rainfall sensor, one can obtain reliable rainfall information in the vicinity of the rainfall sensor. Such rainfall sensor, therefore, provides rainfall information in a localized area rather than just a specific spot. This information is thus much more reliable than the traditional rain sensor. Moreover, unlike conventional rainfall sensors, such a sensing method can provide real-time rainfall information, and it is not affected by the local variation of the rainfall, leaves and debris around the sensor, strong wind and mechanical failures.

[0040] System Description - Rainfall Sensor

[0041] The architecture of the wireless rainfall sensor is shown 20 in Fig. 2. It consists of the following:

[0042] A wireless receiver 21 for receiving cellular signals from multiple base stations across multiple frequency bands. The wireless receiver is interconnected to antennas 22 and a signal digitization unit 24. The antennas 22 can have different polarities and different number of elements. They can use beamforming or just omnidirectional elements.

[0043] The signal digitization unit 24 can be a software defined radios or a customised digital receiver that supports wireless signals in cellular communication frequency bands. It can also be cellular transceiver modules, such Qualcomm Snapdragon X20 LTE Modem [https: / / www.qualcomm.com / products / technology / modems / snapdragon-modems-4g-lte-x20], which can produce typical cellular signal metrics.

[0044] The signal processing unit 24 that processes the above signals to obtain rainfall information. The signal processing unit can consist of digital processor (DSP), field programming gate array (FPGA) or other computing platforms that can execute rainfall sensing algorithms.

[0045] A wireless transmitter 23 to transfer the rainfall information to the Internet. The wireless transmitter 23 consists of microwave transmitter and is connected to output antennas 25. The wireless network can be cellular networks, loT networks or satellite networks.

[0046] The functionalities of the signal processing unit 24 will now be described with reference to Fig. 9.

[0047] Turning now to Fig. 9 there is illustrated a series of steps 90 in the method of an embodiment.

[0048] Step 1 (91): A scheme for cell search and identification. The cell search can be done by surveying the frequency spectrum of cellular communication systems. If the energy at a cellular frequency is greater than the noise power, a cell is deemed to exist at the frequency. The frequency bands used for cellular communications are generally defined in 3GPP specifications and the frequency regulation organisation of the country where the rainfall sensor is deployed.

[0049] At the frequency bands deduced from the above survey, cell identification can be performed by searching for the known primary and secondary synchronization signals used bycellular communication systems. Such synchronization signals are defined in the 3GPP specifications (as per https: / / www.3gpp.oi-g / ). The combination of primary and secondary synchronization signals detected can be used to uniquely deduce a cell identification (an index number, e.g., 5).

[0050] Step 2 (92): Sensing signal extraction and conditioning.

[0051] Based on the cell identification obtained above, the locations of cell reference signals over time and frequency domains can be further deduced. Dividing the received signals at these locations by the transmitted signals (which are defined the 3GPP specification) leads to the channel coefficients at those locations. Interpolation can be performed based on the channel coefficients estimated above to obtain the channel state information over the whole observed time and frequency ranges. Filters can be applied to the estimated channel state information to remove those frequency components that have little contribution to rainfall sensing. Advanced techniques, such as principal component analysis, can also be applied to focus signals onto several principal components

[0052] Step 3 (93): The embodiments can also include a method for fusing the processed channel state information obtained above over multiple base stations and frequency bands. When there is a rainfall around the rainfall sensor, the variation of signals from different base stations and across different frequency bands will be different. By aggregating the signals from different base stations across different frequency bands, one can get much a richer data set for much more reliable rainfall sensing. Possible methods to realise such fusion include but not limited to weighted averaging, principal component analysis, fuzzy logic fusion, Bayesian fusion, Kalman filtering and neuro network filtering etc [4] .

[0053] Step 4 (94): A set of trained machine learning models for converting fused signals obtained above to rainfall information can than be provided. The machine learning models can include recurrent neural networks (RNNs) and their variants like Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), convolutional neural networks and support vector machines etc.

[0054] The rainfall sensor can also learn the deployment environment to select the best suited machine learning model with well-trained parameters for in-situ rainfall sensing.

[0055] Scenario Diagram

[0056] A scenario diagram of the rainfall sensing technique is illustrated 30 in Fig. 3. The wireless rainfall sensor receiver 31 is placed in the middle 32 of a star topology with lines extendingto nearby base station transmitters. These lines denote propagation paths from cellular base stations to the proposed rainfall sensor. The star topology shown in the figure is up to about 800 metres wide.

[0057] As shown Fig. 3, the receivers 31 passively capture the microwave signals from the nearby base stations, which constantly transmit mobile signals. When rainfall happens, the propagation paths between base stations and the receiver will be affected from different perspectives. Through extracting typical signal metrics from the received microwave signals and feeding these features to a trained machine learning model, rainfall intensity can be estimated with high accuracy in real time.

[0058] Exemplary Processing and Results

[0059] Based on the deployment scenario illustrated in Fig. 3, a rainfall sensing procedure can be further explained by illustrating key results during the critical processing steps.

[0060] Data collection: Between 11th Jan 2024 and 21st Jan 2024, mobile signals over 8 carrier frequencies are collected at the central location 32 specified in Fig. 3. The frequencies were: 763MHz, 778MHz, 875MHz, 1812.5MHz, 1825.6MHz, 1840MHz, 1857.5MHz, 2117.5MHz. These are typical LTE mobile carrier frequencies in Sydney, Australia. The digital baseband signals over these frequencies were collected and processed offline for validating the rainfall sensing techniques.

[0061] Step 1: As illustrated 40 in Fig. 4, in Step 1, cell search is performed first by identifying a cell identification number (ID). This is done by matched filtering the synchronisation signals of base stations. These signals are known according to the cellular communication specifications. They are unique under each cell ID. So initially a local signal templates can be created based on all cell IDs and perform matched filtering using these local templates as filtering coefficients. When a peak signal is detected, as shown in Fig. 4, a cell is detected for the carrier frequency with the ID identified, which is the index of the local signal template. For the 8 carrier frequencies mentioned above, their cell IDs are identified, as depicted in Fig. 4, along with their matched filtering results with strong signal peaks. These peaks suggest that the cell is effectively detected and identified.

[0062] Step 2: Turning now to Fig. 5, with the cell IDs successfully detected in Step 1, the locations of reference signals can be decoded based on communication specifications, as stated in Step 2, which then enables us to extract key signal metrics. Through the steps discussed above in respect of step 2, clean CSI signals can be obtained over the time -frequency resources available in the received signals. Then, as stated above, a principal component analysis can be performed over sub-carriers (i.e., frequency-domain resources), which is then converted to the signal-to-interference-plus-noise (SINR) metric in this example. The SINR is cell specific, as it is calculated based on the cells identified from the previous steps and based on the reference signals specific to detected cells. The SINRs of the 8 cells detected are illustrated in Fig. 5. To highlight the impact of rainfall on this key signal metric, the ground-truth rainfall intensity, as presented in millimetre per hour (mm / h) is also plotted for each cell’s SINR curve. A clear impact impact of rainfall on the SINR of a cell based on the variations of SINRs in different rainfall intensities can be seen. Moreover, the diverse reactions of different cellular signals to the rainfall intensities is seen.

[0063] Step 3: The diversities in the results provided above can be joined through further information fusion, as stated in Step 3 above of the invention. Here, to exemplify the impact of cell fusion, a weighted average is performed over the SINRs obtained above. The weights are calculated based on localised signal segments. For each signal point, a window is applied to select signals around it. Here, 160 points are taken on each side of the window. The median is taken to replace the signal. The signal minus the median is larger than one times the standard deviation of all signals in the window. This is based on the Hapel filtering [https : / / an .mathworks . com / help / signal / ref / hampel .html] .

[0064] Fig. 6 shows the fused SINRs over all cells.

[0065] Step 4: The fused signal is then used for machine learning training and testing. Typical time and frequency domain features can be extracted based on the fused signals fortraining machine learning models. Fig. 7 illustrates 11 typical signal features which are often used formachine learning training. In this example, we use these features to train a Gaussian process regression (GPR) model for rainfall estimation. This embodiment is not limited to GPR, and these features can be employed fortraining all regression models in theory.

[0066] Fig. 7 illustrates typical signal features for training machine learning models, which are extracted based on the fused signals as illustrated in Fig. 6.

[0067] For illustration purposes, a rational quadratic GPR is used here, where the basis function is constant, the Kemal function is the rational quadratic function, the isotropic kernel is applied during GPR training and an automatic scaling is performed fortraining. The signal sequences of key signal features obtained above are used for the training. One third of the signals for each feature are randomly selected over the whole time span for training. Then the trained results are used for estimating the rainfall intensity over the whole time span. The ground truth and estimation rainfallintensities are plotted in Fig. 8. We see that the estimated rainfall matches the ground truth accurately, which validates the effectiveness and high performance of the invented rainfall sensing techniques.

[0068] Fig. 8 illustrates the rainfall sensing results employing a GPR model.

[0069] In some embodiments, there is provided an apparatus using cellular signals to measure the rainfall of a local area, including: a wireless receiver for receiving cellular signals from multiple base stations across multiple frequency bands; a signal processing unit that processes the above signals to obtain rainfall information; and a wireless transmitter that transmits the rainfall sensing results to an internet-based data management platform.

[0070] In some embodiments, the cellular signals used can be from either singular or multiple neighbouring base stations. These base stations can be 4G, 5G and 6G. The signals can include synchronization signals, broadcast pilots, cell-specific reference signals, and channel state information signals. The cellular signals can be used from either singular or multiple frequency bands.

[0071] The embodiments also provide for a new method for data fusion and signal processing to obtain rainfall information.

[0072] The wireless transmitter used to report rainfall sensing results can be WiFi, cellular, Bluetooth, LoRaWAN, satellite etc.

[0073] The apparatus of the embodiments can also be used for detecting snowfall, hail and other weather factors having similar impact on microwave signals. In that case, with the overall sensing scheme remained the same, the machine learning models can be adapted to take into account the difference between the impacts of rainfall and snowfall on the propagation of wireless signals.

[0074] The signal processing unit can be replaced with a central control unit which controls the reception of the wireless cellular signals and transmits the baseband signals to a cloud data management and computing platform using the wireless transmitter in Fig. 1. In this case, the signal processing methods and machine learning models in this embodiment are still applicable, except they are performed in an internet-based cloud platform.Interpretation

[0075] Reference throughout this specification to “one embodiment”, “some embodiments” or “an embodiment” means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases “in one embodiment”, “in some embodiments” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.

[0076] As used herein, unless otherwise specified the use of the ordinal adjectives "first", "second", "third", etc., to describe a common object, merely indicate that different instances of like objects are being referred to, and are not intended to imply that the objects so described must be in a given sequence, either temporally, spatially, in ranking, or in any other manner.

[0077] In the claims below and the description herein, any one of the terms comprising, comprised of or which comprises is an open term that means including at least the elements / features that follow, but not excluding others. Thus, the term comprising, when used in the claims, should not be interpreted as being limitative to the means or elements or steps listed thereafter. For example, the scope of the expression a device comprising A and B should not be limited to devices consisting only of elements A and B. Any one of the terms including or which includes or that includes as used herein is also an open term that also means including at least the elements / features that follow the term, but not excluding others. Thus, including is synonymous with and means comprising.

[0078] As used herein, the term “exemplary” is used in the sense of providing examples, as opposed to indicating quality. That is, an “exemplary embodiment” is an embodiment provided as an example, as opposed to necessarily being an embodiment of exemplary quality.

[0079] It should be appreciated that in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of this invention.

[0080] Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those skilled in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0081] Furthermore, some of the embodiments are described herein as a method or combination of elements of a method that can be implemented by a processor of a computer system or by other means of carrying out the function. Thus, a processor with the necessary instructions for carrying out such a method or element of a method forms a means for carrying out the method or element of a method. Furthermore, an element described herein of an apparatus embodiment is an example of a means for carrying out the function performed by the element for the purpose of carrying out the invention.

[0082] In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

[0083] Similarly, it is to be noticed that the term coupled, when used in the claims, should not be interpreted as being limited to direct connections only. The terms "coupled" and "connected," along with their derivatives, may be used. It should be understood that these terms are not intended as synonyms for each other. Thus, the scope of the expression a device A coupled to a device B should not be limited to devices or systems wherein an output of device A is directly connected to an input of device B. It means that there exists a path between an output of A and an input of B which may be a path including other devices or means. "Coupled" may mean that two or more elements are either in direct physical or electrical contact, or that two or more elements are not in direct contact with each other but yet still co-operate or interact with each other.

[0084] Thus, while there has been described what are believed to be the preferred embodiments of the invention, those skilled in the art will recognize that other and further modifications may be made thereto without departing from the spirit of the invention, and it is intended to claim all such changes and modifications as falling within the scope of the invention. For example, any formulas given above are merely representative of procedures that may be used. Functionality may be added or deleted from the block diagrams and operations may be interchanged among functional blocks. Steps may be added or deleted to methods described within the scope of the present invention.

Claims

CLAIMS:

1. An apparatus for measuring rainfall intensity in a local geographic area, the apparatus including: an initial learning system including: a wireless receiver for receiving a series of cellular signals emitted from multiple base station signals in the neighbourhood of the wireless receiver; a monitoring unit for monitoring the rainfall in the region of the wireless receiver; a machine learning unit for processing the correlation between the rainfall in a region and the corresponding attenuation of the cellular signals of multiple base stations for an extended period of time to produce a subsequent prediction model; and subsequent reporting means for reporting on a current correlation between the series of cellular signals and expected rainfall.

2. A method of monitoring the rainfall over an extended period of time, the method including the steps of:(a) initially monitoring a plurality of cellular base station emissions to identify each base station location and signal characteristics;(b) extracting frequency and signal characteristics from the base station emissions, in conjunction with rainfall reports for the area around the monitoring system, utilising machine learning to produce an extended correlation model of the change in signal characteristics with respect to rainfall; and(c) subsequently utilising the extended correlation model to output a rainfall prediction based on the received frequency and signal characteristics.

3. An apparatus using cellular signals to measure the rainfall of a local area, including: a wireless receiver for receiving cellular signals from multiple base stations across multiple frequency bands; a signal processing unit that processes the above signals to obtain rainfall information; and a wireless transmitter that transmits the rainfall sensing results to an internet-based data management platform.

4. An apparatus as claimed in claim 3 wherein cellular signals used can be from either singular or multiple neighbouring base stations.

5. An apparatus as claimed in claim 4 where the base stations are 4G, 5G or 6G.

6. An apparatus as claimed in claim 3 where the signals include one of synchronization signals, broadcast pilots, cell-specific reference signals, and channel state information signals.

7. An apparatus as claimed in claim 3 wherein the cellular signals are used from either singular or multiple frequency bands.

8. An apparatus as claimed in claim 3 wherein the wireless transmitter used to report rainfall sensing results can utilise WiFi, cellular, Bluetooth, LoRaWAN, or satellite.

9. An apparatus as claimed in any previous claim wherein it it suited to detect at least one of snowfall, hail and other weather factors having an impact on microwave signals.

10. An apparatus as claimed in any previous claim where the signal processing unit is replaced with a central control unit which controls the reception of the wireless cellular signals and transmits the baseband signals to a cloud data management and computing platform using a wireless transmitter for derivation of a machine learning model.

Images