CLIAMS:1. A computer implemented system for converting low resolution data to high resolution data, said system comprising:
i. a first extractor module adapted to extract samples of high resolution data signals from a physical environment;
ii. a second extractor module adapted to extract samples of low resolution data signals from a data source;
iii. a first transceiver adapted to receive and transmit the high resolution data signal samples extracted by the first extractor module;
iv. a second transceiver adapted to receive and transmit the low resolution data signal samples extracted by the second extractor module;
v. a sparsity checking module adapted to receive the extracted high resolution data signal samples from the first transceiver, said sparsity checking module comprising:
• a domain examining module adapted to examine the domain in which the extracted high resolution data signal samples are sparse and also adapted to establish a sparse domain matrix;
and
vi. a compressed sensing module adapted to receive the sparse domain matrix from the domain examining module and the low resolution data signal samples from the second transceiver, said compressed sensing module comprising:
• a sensing matrix designing module configured to design a sensing matrix based on the relationship between the low resolution data signal samples and the expected high resolution data signal;
• a multiplier module adapted to receive the designed sensing matrix from the sensing matrix designing module and the sparse domain matrix from the domain examining module, and further adapted to multiply elements of the sparse domain matrix with elements of the designed sensing matrix to obtain a sensing basis;
• a reconstruction module adapted to receive the sensing basis from the multiplier module and the low resolution data signal samples from the second transceiver, and further adapted to process the received data signals simultaneously to obtain high resolution data signal
2. The system as claimed in claim 1, wherein said data source is a smart meter selected from the group consisting of an electricity meter, a gas meter, a water meter and a heat meter.
3. The system as claimed in claim 1, wherein said second extractor module extracts low resolution data signal samples from the data source by sampling methods selected from the group consisting of decimation sampling method and random sampling method.
4. The system as claimed in claim 1, wherein said domain examining module examines, the domain in which the extracted high resolution data signal samples are sparse, by methods selected from the group consisting of Noiselets, Curvelets, Discrete Fourier Transform, Discrete Cosine Transform, Discrete Wavelet Transform and dictionary learning methods.
5. The system as claimed in claim 1, wherein said reconstruction module processes the received data signals by using reconstruction methods selected from the group consisting of L-1 Magic methods and pursuit methods.
6. A computer implemented method for converting low resolution data to high resolution data, said method comprising the steps of:
• extracting a plurality of high resolution data signal samples from a physical environment;
• transmitting the extracted high resolution data signal samples;
• extracting a plurality of low resolution data signal samples from a data source;
• transmitting the extracted low resolution data signal samples;
• receiving the extracted high resolution data signal samples and examining a domain in which the extracted high resolution data signal samples are sparse;
• establishing and transmitting a sparse domain matrix;
• designing a sensing matrix based on the relationship between the low resolution data signal samples and the expected high resolution data signal;
• receiving the designed sensing matrix and the sparse domain matrix and multiplying elements of the received designed sensing matrix with elements of the received sparse domain matrix to obtain a sensing basis; and
• receiving the sensing basis and the low resolution data signal samples and processing the received sensing basis and the received low resolution data signal samples simultaneously to obtain high resolution data signal.
7. The method as claimed in claim 6, wherein the step of extracting the low resolution data signal samples from the data source includes a step of sampling by sampling methods selected from the group consisting of decimation sampling method and random sampling method.
8. The method as claimed in claim 6, wherein the domain in which the extracted high resolution data signal samples are sparse involves step of examining by methods selected from the group consisting of Noiselets, Curvelets, Discrete Fourier Transform, Discrete Cosine Transform, Discrete Wavelet Transform and dictionary learning methods.
9. The method as claimed in claim 6, wherein the step of processing the received data signals includes step of using reconstruction methods selected from the group consisting of L-1 Magic methods and pursuit methods. ,TagSPECI:FIELD OF THE DISCLOSURE
The present disclosure relates to the field of converting resolution of data signals.
DEFINITIONS OF TERMS USED IN THE SPECIFICATION
The expression ‘high resolution’ used hereinafter in this specification refers to finely detailed data that can be reconstructed accurately.
The expression ‘physical environment’ used hereinafter in this specification refers to an area including houses and buildings from where data measurements corresponding to consumption of household utilities that are particularly used by humans is obtained.
The expression ‘data source’ used hereinafter in this specification refers to sources including smart meters and sensors that are used to record consumption of household utilities particularly used by humans.
The expression ‘sparse domain’ used hereinafter in this specification refers to domain in which signals can be efficiently represented.
BACKGROUND
The advancement in the networking technologies in the last decade have provided reliable connections to disparate devices in the homes and offices, supported metering of other utilities such as gas and water. Monitoring and synchronization of wide area networks expand the horizon of research in this technology domain by using prototype sensors which are capable of providing rapid analysis of anomalies in electricity quality over a large geographical area. However, in current scenario a signal from a data source like a smart meter is sampled at rates either too low to be used for applications like activity monitoring or too high to be transmitted to remote locations for monitoring purposes. This is insufficient for tasks that involve activity monitoring which require continuous transmission updates and monitoring data.
Further the current approaches for activity monitoring are not generic to be applied to smart meters across various geographical locations. Therefore, there is a long felt need for a system and method to process meter data and to convert the low resolution data in to high resolution data.
OBJECTS
An object of the present disclosure is to obtain high-resolution data signal from low-sampled data signal.
Another object of the present disclosure is to utilize existing compressed sensing techniques to obtain a high resolution data signal.
Yet another object of the present invention is to provide interpolation to obtain finer resolution of the low-sampled data signal using compress-sensing techniques.
Other objects and advantages of the present disclosure will be more apparent from the following description when read in conjunction with the accompanying figures, which are not intended to limit the scope of the present disclosure.
SUMMARY
The present disclosure envisages a computer implemented system and method for converting low resolution data to high resolution data.
Typically, in accordance with the present disclosure the system for converting low resolution data to high resolution data comprises:
i. a first extractor module adapted to extract samples of high resolution data signals from a physical environment;
ii. a second extractor module adapted to extract samples of low resolution data signals from a data source;
iii. a first transceiver adapted to receive and transmit the high resolution data signal samples extracted by the first extractor module;
iv. a second transceiver adapted to receive and transmit the low resolution data signal samples extracted by the second extractor module;
v. a sparsity checking module adapted to receive the extracted high resolution data signal samples from the first transceiver, the sparsity checking module comprising:
• a domain examining module adapted to examine a domain in which the extracted high resolution data signal samples are sparse and also adapted to establish a sparse domain matrix; and
vi. a compressed sensing module adapted to receive the sparse domain matrix from the domain examining module and the low resolution data signal samples from the second transceiver, the compressed sensing module comprising:
• a sensing matrix designing module configured to design a sensing matrix based on the relationship between the low resolution data signal samples and the expected high resolution data signal;
• a multiplier module adapted to receive the designed sensing matrix from the sensing matrix designing module and the sparse domain matrix from the domain examining module, and further adapted to multiply elements of the sparse domain matrix with elements of the designed sensing matrix to obtain a sensing basis;
• a reconstruction module adapted to receive the sensing basis from the multiplier module and the low resolution data signal samples from the second transceiver, and further adapted to process the received data signals simultaneously to obtain high resolution data signal
In accordance with the present disclosure, there is provided a method for converting low resolution data to high resolution data, the method includes following steps of:
• extracting a plurality of high resolution data signal samples from a physical environment;
• transmitting the extracted high resolution data signal samples;
• extracting a plurality of low resolution data signal samples from a data source;
• transmitting the extracted low resolution data signal samples;
• receiving the extracted high resolution data signal samples and examining a domain in which the extracted high resolution data signal samples are sparse;
• establishing and transmitting a sparse domain matrix;
• designing a sensing matrix based on relationship between the low resolution data signal samples and the expected high resolution data signal;
• receiving the designed sensing matrix and the sparse domain matrix and multiplying elements of the designed sensing matrix with elements of the sparse domain matrix to obtain a sensing basis; and
• receiving the sensing basis and the low resolution data signal samples and processing the received sensing basis and the received low resolution data signal samples simultaneously to obtain high resolution data signal.
BRIEF DESCRIPTION OF ACCOMPANYING DRAWINGS
The system of the present disclosure will now be described with the help of the accompanying drawings, in which:
FIGURE 1 illustrates the schematic of the system that converts low resolution data to high resolution data.
DETAILED DESCRIPTION OF THE ACCOMPANYING DRAWINGS
A preferred embodiment of a system and method of the present disclosure will now be described in detail with reference to the accompanying drawings. The preferred embodiment does not limit the scope and ambit of the disclosure. The description provided is purely by way of example and illustration.
The embodiments herein and the various features and advantageous details thereof are explained with reference to the non-limiting embodiments in the following description. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
The system of the present disclosure examines data from a data source, as a set of measurements, in different domains to exhibit maximum sparsity in order to select a basis function. The selected basis function along with the set of measurements result in interpolated original length of data that can be used for further analysis.
Referring to the accompanying drawings, Figure 1 illustrates the system 100 for converting low resolution data to high resolution data. In one embodiment of the present disclosure, a smart meter is used as a data source from which the data signals are sampled. This sampled data is considered to be a decimated version of the desired high resolution data.
The system 100 includes a first extractor module 102 that extracts high resolution data signal samples from a physical environment such as buildings, houses and the like. The system 100 also includes a second extractor module 110 that extracts low resolution data signal samples from a data source. In accordance with one embodiment, the extraction of high resolution data signal samples from the physical environment includes extraction from smart meters and/or sensors installed in physical environments. In accordance with one embodiment, the data source is a smart meter and the second extractor module 110 uses sampling techniques including but not limited to decimation or random sampling to obtain low resolution data signal samples. The method of decimation or down sampling a signal is a process of reducing the sampling rate of the signal whereas, in random sampling each individual sample is chosen randomly such that each sample has same probability of being chosen at any stage during the sampling process. These techniques are used to obtain low resolution data signal samples that are sub-sampled versions of a higher resolution smart meter data. A first transceiver 104 receives the extracted data from the first extractor module 102. The first transceiver 104 transmits the received high resolution data signal samples to a sparsity checking module 106. The sparsity checking module 106 comprises a domain examining module 108 that examines a domain in which the extracted high resolution data signal samples are sparse. The high resolution data signal samples are checked for sparsity, wherein the methods for checking sparsity include but are not limited to Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and the like. This sparsity is determined by checking the number of transform coefficients required to reconstruct the data. The process of checking the number of transform coefficients is typically an offline process and is carried out only once for a particular data source within the physical environment. This process provides behavior of measurements from that particular data source. On completion of sparsity determination, the checking process need not be repeated for that particular data source. Upon examination of the sparse domain by the domain examining module 108, information related to a sparse domain matrix is transmitted to a compressed sensing module 114. The system 100 includes a second transceiver 112 that receives the extracted data from the second extractor module 110. The low resolution data signal samples extracted by the second extractor module 110 and transmitted by the second transceiver 112 are received by the compressed sensing module 114. These low resolution data signal samples are further processed by a reconstruction module 120 that is configured within the compressed sensing module 114.
In accordance with an exemplary embodiment, the low resolution data signal samples are treated as compressed sensed (CS) measurements, 'y', and a higher resolution data signal 'x' is to be obtained by the system. Following equation (1) shows the relation between ‘x’ and ‘y’:
y = Ф*x (1)
where, it is assumed that ‘y’ is a low-sampled version of ‘x’, either through decimation of ‘x’ or through random selection of a few samples of ‘x’, and ‘Ф’ is a sensing matrix. The compressed sensing module 114 comprises a sensing matrix designing module 116 that designs the sensing matrix ‘Ф’ to satisfy the relation between ‘y’ and ‘x’ based on the equation (1). In accordance with an embodiment, the sensing matrix ‘Ф’ designed by the sensing matrix designing module 116 can be an M*N matrix (where length of M is less than length of N), wherein for each sample considered form the high resolution data, the corresponding element in the matrix is a ‘1’ and for the discarded samples, the corresponding element in the matrix are zeros. i.e. if M=288 and N=1440, and if a sample (for e.g. 5th sample) is non-zero in the 288-length data, then the 5th element in the 5th row will be ‘1’, rest of the elements in the row will be zero.
Additionally, the higher resolution data signal ‘x’ from equation (1) is obtained by the following equation (2):
x = φ*s (2)
where, ‘φ’ is a sparse domain matrix established by the domain examining 108 on the basis of sparsity determined by the sparsity checking module 106 and ‘s’ refers to sparse domain coefficients that contain information about ‘x’ in ‘φ’ domain. The data source is examined offline to determine the sparsity.
The compressed sensing module 114 further comprises a multiplier module 118 that receives the sensing matrix ‘Ф’ from the sensing matrix designing module 116 and, the sparse domain matrix ‘φ’ on determination of sparsity domain from the domain examining module 108. The sparse domain matrix information is sent only once during the initial setup phase. The multiplier module 118 multiplies elements of the sparse domain matrix φ with elements of the sensing matrix Ф to obtain a sensing basis ‘A’ (A is termed 'sensing basis' since Ф is a sensing matrix and φ is a sparsity basis and A is the combination of the two). This is denoted in following equation (3):
A = Ф * φ (3)
This sensing basis ‘A’ is then utilized by the reconstruction module 120 that cooperates with the compressed sensing module 114. The reconstruction module 120 receives the sensing basis from the multiplier module 118 and the low resolution data signal samples from the second transceiver 112, and processes the received data signals simultaneously to obtain high resolution data signal. The reconstruction module 120 uses reconstruction methods including but not limited to L-1 Magic and similar pursuit methods. L-1 Magic is used for solving convex optimization, central to compressive sampling. The output of the reconstruction module 120 is desired high resolution data of the data source as denoted by ‘x’ in equation (1).
The aforementioned embodiment deals with the data source measurements that are assumed due to decimation or random sampling of a higher resolution data. In another embodiment, the compressed sensed (CS) measurements ‘y’ is calculated by equation (4) as follows:
y = (Ф2 Φ1) x (4)
where, the overall measurement matrix Ф is equal to Ф2 Φ1 and is random. Φ. can be random or deterministic. As in the previous embodiment, ‘x’ is the high-resolution version of the data.
The low resolution measurements from the data source y. = Φ1 x with Φ1 capturing the decimation process, is further operated by Ф2 in the reconstruction module 120 to result in y = (Ф2 Φ1) x. The resultant 'y' is fed to the reconstruction module 120 and processed by reconstruction methods including but not limited to L-1 Magic and similar pursuit methods.
Ф2 reduces the size of ‘y’ and is an appropriately chosen random matrix. The procedure ensures reconstruction of ‘x’ from ‘y’ even in the cases where strict sparsity is not exhibited by ‘x’.
Using the linearity property of the operations the matrix multiplication is delegated with Ф2 to the reconstruction module 120 at the cost of transmitting or capturing slightly more measurements from the data source. This leads to using the data source measurements without any add-ons while delegating the computation to the receiver or reconstruction module, using the above mentioned mathematical formulation.
In accordance with another embodiment, distribution-based sampling is introduced to obtain ‘y’ in order to overcome the limitation of uniformly spaced low-resolution samples that may not be conducive to allow good reconstruction.
For this, samples are picked effectively at the receiver side using a probability distribution which in turn defines ‘Ф2’. Such distribution based on random sampling generates accurate reconstruction. All these computations are carried out in the reconstruction module, without any add-on to the data source.
Thus, the present disclosure facilitates use of data signals from a data source to establish a sparse domain for the data signals by studying behavior of the data signals in various domains in order to obtain precise high resolution data.
TECHNICAL ADVANCEMENTS
A computer implemented system and method for converting low resolution data to high resolution data in accordance with the present disclosure described herein above has several technical advancements including but not limited to the realization of:
• a system that obtains high-resolution data signal from low-sampled data signal;
• a system that utilizes existing compressed sensing techniques to obtain a high resolution data signal; and
• a system that provides interpolation to obtain finer resolution of the low-sampled data signal using compress-sensing techniques.

The foregoing description of the specific embodiments will so fully reveal the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the spirit and scope of the embodiments as described herein.