TY - JOUR
AU - Wang,, Lei
AB - Abstract Building Energy Internet of Things could collect and analyse various types of building energy consumption data in real time by means of low-energy consumption and high-precision sensing technology. In this paper, a low-energy consumption data transmission and fusion algorithm SMART-RR (Slice Mix Agg RegaTe-Repeatablibity Reduction) is proposed. Taking advantage of the periodic repeatability and data redundancy of building energy consumption data, a data fusion strategy with unequal long time intervals and adding repeatability reduction factor is proposed. The simulation results show that SMART-RR algorithm is a low-energy data transmission and fusion algorithm with small data traffic, high privacy protection and high accuracy. 1 INTRODUCTION Building energy consumption monitoring is one of the research hotspots in the field of modern urban energy management. Building energy consumption monitoring system is used for real-time monitoring of building energy consumption, including electricity consumption, heat consumption, water consumption and other modes of energy consumption [1]. It can effectively reflect the building energy consumption, and provide data basis for urban energy management and energy control [2]. The traditional energy consumption monitoring data acquisition and transmission mode relying on industrial bus and industrial Ethernet is limited by wired network, and is restricted in wiring complexity, system scalability and maintainability [3]. In recent years, the concept of Building Energy Internet of Things (BEIoT) has been proposed and attracted the attention of relevant scholars [4]. BEIoT relies on the theory and technology of wireless sensor network (WSN) and building energy management to lay the wireless sensors in or around the target buildings that need to be monitored and form a multi-layer architecture including data acquisition layer, data transmission layer and data analysis and control layer by means of sensor autonomous networking [5]. BEIoT collects energy consumption data of buildings and surrounding environment through various types of sensors, transmits them through wireless network and collects them to the central control node for data analysis, which provides a basis for energy consumption analysis and energy equipment control. Building energy consumption data acquisition system based on BEIoT can collect multi-modal building energy consumption data in real time, which provides a means for fast and accurate energy consumption analysis and energy control. Researchers have conducted in-depth research on BEIoT data fusion algorithm [6–8]. The building energy consumption data in BEIoT are uploaded and fused step by step along the wireless communication network. These data contain a lot of privacy information related to building attributes and energy consumption. Once the nodes in the sensor network are maliciously invaded, the sensitive privacy information will be monitored, which will lead to abnormal building energy consumption data fusion and result in data security disaster of building energy management and user trust in building energy consumption data monitoring based on BEIoT. In order to improve the security of WSN data transmission and fusion, researchers have proposed a variety of encryption algorithms, such as key distribution algorithm, homomorphic encryption algorithm, etc. [9, 10]. Key encryption algorithm protects sensor nodes in WSN by pre-distributing key to prevent sensitive information from being leaked during network transmission. However, due to frequent key distribution and decryption operations in data fusion process, data fusion efficiency is often reduced and energy consumption of sensor nodes is increased. In order to improve the efficiency of data fusion, a WSN-oriented homomorphic encryption algorithm is proposed, which performs fusion computation based on encrypted ciphertext [11]. However, the privacy protection of this method is weakened. If any nodes in sensor networks are maliciously intruded, the privacy data in adjacent nodes may not be effectively protected. Madden et al. [12] presented the Tiny AGgregation (TAG) service for aggregation in low-power, distributed, wireless environments, which allowed users to express simple, declarative queries and have them distributed and executed efficiently in networks of low-power, wireless sensors. Cluster-based Private Data Aggregation and Slice Mix Agg RegaTe (SMART) are proposed based on the key distribution algorithm [13, 14]. After collecting data, sensor nodes first perform fragmentation operations and transmit them to different nodes. In this way, the privacy protection of sensitive data is enhanced. Even if some nodes in the network are intruded, they could also be clearly resolved to complete and accurate data. After analysing the shortcomings of the SMART algorithm that the accuracy of data fusion decreases due to the fragmentation of the transmitted data, a series of improved algorithms with multiple optimization factors are proposed [15]. The performance of the proposed algorithm in privacy protection, accuracy and fragmentation reception rate is verified by comparison experiments and analysis. Because the building energy consumption data collected by BEIoT often have multi-modal characteristics, data fragmentation strategy for multi-modal health data and data fusion algorithm based on unequal time interval data can be further studied on the basis of existing privacy protection algorithms, so as to improve the privacy protection of health data and reduce the network traffic of BEIoT [16]. Aiming at the security, low power consumption and accuracy requirement of data transmission and fusion for building energy consumption monitoring, this paper introduces local optimization factor based on SMART and related algorithms, and proposes a low-power data transmission and fusion algorithm, which can improve the accuracy of building energy consumption monitoring data on the premise of reducing the redundancy of sensor node data and the traffic of sensor network in BEIoT. Section 2 discusses the building energy consumption monitoring system based on WSN, and introduces the topological structure of the building energy consumption monitoring system and low-energy consumption sensor nodes. Section 3 proposes a data fusion model based on multi-hop self-organizing network and a data fusion algorithm with optimization factor from the defect analysis of the application of SMART algorithm in BEIoT. Section 4 analyses and evaluates the performance of the algorithm in terms of data traffic, data accuracy and privacy protection through simulation experiments. Finally, Section 5 proposes a brief summary and looks forward to the future work. 2 BUILDING ENERGY CONSUMPTION DATA MONITORING BEIoT locates many types of low-power and high-precision sensors in and around the monitored buildings. With the help of wireless sensor technology, it forms a WSN structure that can collect and transmit real-time information of building energy consumption, and supports building energy consumption monitoring and data analysis. This section first introduces the topology of building energy consumption monitoring system based on WSN, and then describes the low-energy consumption sensor nodes for building energy consumption data acquisition and transmission. 2.1 Topology of BEIoT The topological structure of building energy consumption data monitoring system based on WSN is shown in Figure 1, which is realized by networking of a variety of data acquisition sensor nodes, data transmission nodes and data analysis nodes inside and around buildings. Figure 1 Open in new tabDownload slide Structural sketch of building energy consumption monitoring system based on IoT. Figure 1 Open in new tabDownload slide Structural sketch of building energy consumption monitoring system based on IoT. Multi-type data acquisition sensor nodes located inside and around buildings acquire multi-modal energy consumption and environmental data such as electricity consumption, water consumption, geothermal information, temperature information, humidity information of the monitored buildings, and transmit them periodically to data transmission nodes in the cloud through wireless communication network. After collecting multi-modal building energy consumption and environmental data, the data transmission node carries out preliminary noise filtering, and carries out data fusion calculation at the first level according to the data type, and uploads the calculation results to the data fusion and analysis node. The data fusion and analysis node obtains the preliminary processed building energy consumption information, carries on the final data fusion and calculation, obtains the results of building energy consumption data analysis and environmental assessment indicators and assists municipal administrators and engineering technicians in building energy consumption monitoring, energy consumption evaluation and environmental assessment. 2.2 Sensor nodes for building energy consumption monitoring In the practical application of wireless data acquisition and transmission in the field of building energy consumption monitoring and environmental monitoring, low-energy consumption and high-precision wireless sensor nodes are often used to locate and collect data. Firstly, wireless sensor nodes used to collect building energy consumption and environmental information are limited by energy, so energy consumption of sensor nodes must be strictly controlled to prolong the life cycle of BEIoT. The energy of wireless sensor nodes in BEIoT is mainly consumed in the process of wireless network communication, so it is necessary to minimize the data traffic and data redundancy between sensor nodes. Secondly, due to the influence of external environment, electromagnetic interference and other factors, the data collected by BEIoT sensor nodes are often mixed with a variety of noise signals, which reduces the accuracy and integrity of building energy consumption data. Therefore, wireless sensor nodes with lower energy consumption and higher accuracy are often needed for wireless acquisition and transmission in the field of building energy consumption monitoring and environmental monitoring. In addition, a more efficient and low-power data transmission and fusion method is adopted. 3 LOW-POWER DATA TRANSFER AND FUSION ALGORITHM After acquiring real-time data of building energy consumption, various sensor nodes in BEIoT upload and fuse layer by layer through tree WSN. The data transmission traffic and encryption and decryption mechanism directly affect the energy consumption of BEIoT and the privacy protection of building energy consumption and environmental monitoring data. In this section, BEIoT is abstracted into a tree network with three layers. A data fusion algorithm based on repeatability protocol factor is proposed, and the key distribution mechanism is used to improve the data privacy protection. 3.1 Data fusion model based on multi-hop network structure BEIoT can be represented by a connected digraph |$G(V,E)$|⁠, where vertex |$v(v\in V)$| represents nodes in BEIoT and directed arc |$e(e\in E)$| represents data transmission links between nodes. Typical BEIoT generally consists of three types of nodes: (1) leaf nodes, which are composed of sensors that collect and transmit building energy consumption and environmental information; (2) fusion nodes, which collect building energy consumption data and perform data fusion and calculation; and (3) Query Server (QS) nodes, which are responsible for the final fusion and analysis of building energy consumption data. Three types of nodes constitute a tree structure, among which QS node is the root node to get the data fusion results and provide the basis for further analysis of building energy consumption and environment. The fusion node is responsible for receiving data from leaf node and transferring the fusion calculation to root node. The leaf node collects various types of building energy consumption data and uploads them to the corresponding node. The data fusion process based on tree structure is one-way transmission. The data fusion function structure currently used in WSN can be expressed as formula (1): $$ f(t)=\varphi \left({d}_1(t),\textsf{K} \kern0.5em ,{d}_n(t)\right) $$ (1) Among them, |${d}_i(t),\kern0.5em (i=1,2,\textsf{K}\kern0.5em ,n)$| represents the data collected by node |$i$| at the point of time |$t$|⁠, and operator |$\varphi$| represents the fusion calculation factors, such as count, average, max, min and other sum functions. As a specific application of WSN in the field of building energy and environment monitoring, BEIoT collects and transmits information on building energy consumption and environment of various types and structures. Some types of data have obvious periodic characteristics, and the data of sudden increase or abnormal fluctuation periods are often more valuable for building energy consumption analysis. Therefore, the characteristics of the above fields can be considered in the design of BEIoT data fragmentation and transmission strategy, and a large number of periodic and redundant building energy consumption data can be compressed, focusing on the collection of occasional abnormal data, so as to reduce the data traffic of WSNs. Data fusion function based on unequal time intervals presents formula (2): $$ f\left(\hat{t}\right)=\varphi \left({d}_1\left(\varDelta{t}_1\right),\textsf{K} \kern0.5em ,{d}_n\left(\varDelta{t}_n\right)\right) $$ (2) Among them, |${d}_i(\varDelta{t}_i)\kern0.5em (i=1,2,\textsf{K} \kern0.5em ,n)$| denotes the data collected by node |$i$| within the time interval |$\varDelta{t}_i$|⁠, and |$\hat{t}=[\varDelta{t}_1,\textsf{K} \kern0.5em ,\varDelta{t}_n]$| denotes the minimum common-fold time interval for all nodes. For periodic redundant data, the time interval can be set relatively long, and some duplicate building energy consumption data can be removed, which can reduce data traffic and improve data fusion efficiency. 3.2 Data fusion algorithm based on repetitive reduction factor In this paper, a data fusion algorithm SMART-RR based on Repeatability Reduction is proposed for BEIoT applications. Data encryption and decryption adopt random key distribution strategy. First, a key pool containing |$K$| keys is generated, from which |$k$| keys (⁠|$k<K$|⁠) are randomly selected. Then, the nodes in BEIoT send messages to each other to determine which nodes are assigned the same key, and it is believed that a data transmission link can be established between the nodes with the same key. If the parent and child nodes in the tree structure do not allocate the same key, the transmission link can be established by hop-by-hop mode. It can be seen from the key distribution strategy that if the listener adopts the same key distribution scheme, the probability of data eavesdropping in BEIoT is |$p=k\kern0.5em /\kern0.5em K$|⁠. Generally, the number of keys in the key pool is relatively small, which reduces the probability of data eavesdropping in WSNs. The basic structure of SMART algorithm consists of three stages. (1) Fragmentation stage. Each node divides the collected data into |$J$| slices, in which |$J-1$| slices are sent to |$J-1$| nodes randomly selected from adjacent nodes. (2) Decryption stage. Nodes receive fragmented data and decrypt it with shared key. (3) Fusion stage. Data fusion using TAG algorithm. Some modal data in BEIoT have obvious periodic repeatability. SMART-RR algorithm with repeatability reduction factor divides the time interval of data acquisition and transmission into different time slices according to different modal data, and reduces the periodic repeatability data in time slices to limit the data traffic in order to reduce the energy consumption of WSNs. According to the periodic repeatability of the collected data, nodes can be divided into periodic node set |${V}_r$| (include thermometer, hygrometer and so on) and non-periodic node set |${V}_u$| (include coulombmeter, water flow meter, geothermal meter and so on). The SMART algorithm flow based on repeatability specification factor is as follows. Algorithm: SMART-RR Input: topological network architecture |$G(V,E)$| of wireless sensor network BEIoT, node set |$i\in{V}_r$| and |$i\in{V}_u$|⁠. Output: Data fusion result |$D$|⁠. 1. WHILE|$\kern0.5em i\in V,\kern1em {V}_i\subset V$|DO 2. choose node set |${V}_i$| in |$h$|-hop randomly 3. calculate the data transmission time interval parameter |$\varDelta{t}_i$| of node |$i$| 4. END WHILE 5. IF|$i\in{V}_r$| 6. set time interval |$\varDelta{t}_i=10{t}_r$| 7. IF the time domain characteristic |${F}_{t- domain}$| of 10 cycles equal each other 8. remove repeatability and retain vector 9. ELSE set time interval |$\varDelta{t}_i=\overline{\varDelta t}$| 10. END IF 11. END IF 12. WHILE|${j}_d\in{V}_i$|DO 13. |$J-1$| compressed data sent to node |${j}_d$| 14. |${j}_d$| decryption with shared key 15. END WHILE 16. output result |$D$| Algorithm: SMART-RR Input: topological network architecture |$G(V,E)$| of wireless sensor network BEIoT, node set |$i\in{V}_r$| and |$i\in{V}_u$|⁠. Output: Data fusion result |$D$|⁠. 1. WHILE|$\kern0.5em i\in V,\kern1em {V}_i\subset V$|DO 2. choose node set |${V}_i$| in |$h$|-hop randomly 3. calculate the data transmission time interval parameter |$\varDelta{t}_i$| of node |$i$| 4. END WHILE 5. IF|$i\in{V}_r$| 6. set time interval |$\varDelta{t}_i=10{t}_r$| 7. IF the time domain characteristic |${F}_{t- domain}$| of 10 cycles equal each other 8. remove repeatability and retain vector 9. ELSE set time interval |$\varDelta{t}_i=\overline{\varDelta t}$| 10. END IF 11. END IF 12. WHILE|${j}_d\in{V}_i$|DO 13. |$J-1$| compressed data sent to node |${j}_d$| 14. |${j}_d$| decryption with shared key 15. END WHILE 16. output result |$D$| Open in new tab Algorithm: SMART-RR Input: topological network architecture |$G(V,E)$| of wireless sensor network BEIoT, node set |$i\in{V}_r$| and |$i\in{V}_u$|⁠. Output: Data fusion result |$D$|⁠. 1. WHILE|$\kern0.5em i\in V,\kern1em {V}_i\subset V$|DO 2. choose node set |${V}_i$| in |$h$|-hop randomly 3. calculate the data transmission time interval parameter |$\varDelta{t}_i$| of node |$i$| 4. END WHILE 5. IF|$i\in{V}_r$| 6. set time interval |$\varDelta{t}_i=10{t}_r$| 7. IF the time domain characteristic |${F}_{t- domain}$| of 10 cycles equal each other 8. remove repeatability and retain vector 9. ELSE set time interval |$\varDelta{t}_i=\overline{\varDelta t}$| 10. END IF 11. END IF 12. WHILE|${j}_d\in{V}_i$|DO 13. |$J-1$| compressed data sent to node |${j}_d$| 14. |${j}_d$| decryption with shared key 15. END WHILE 16. output result |$D$| Algorithm: SMART-RR Input: topological network architecture |$G(V,E)$| of wireless sensor network BEIoT, node set |$i\in{V}_r$| and |$i\in{V}_u$|⁠. Output: Data fusion result |$D$|⁠. 1. WHILE|$\kern0.5em i\in V,\kern1em {V}_i\subset V$|DO 2. choose node set |${V}_i$| in |$h$|-hop randomly 3. calculate the data transmission time interval parameter |$\varDelta{t}_i$| of node |$i$| 4. END WHILE 5. IF|$i\in{V}_r$| 6. set time interval |$\varDelta{t}_i=10{t}_r$| 7. IF the time domain characteristic |${F}_{t- domain}$| of 10 cycles equal each other 8. remove repeatability and retain vector 9. ELSE set time interval |$\varDelta{t}_i=\overline{\varDelta t}$| 10. END IF 11. END IF 12. WHILE|${j}_d\in{V}_i$|DO 13. |$J-1$| compressed data sent to node |${j}_d$| 14. |${j}_d$| decryption with shared key 15. END WHILE 16. output result |$D$| Open in new tab Compared with SMART algorithm, SMART-RR algorithm with repetitive protocol factor can greatly reduce the communication load caused by data transmission of building energy consumption monitoring due to periodic repetition. Although each node consumes a certain amount of time and processing resources compared with SMART algorithm in calculating the time interval |$\varDelta{t}_i$| of data transmission, the analysis shows that the time complexity is |$O(1)$| and does not need to be calculated every time. The extra cost can be neglected. 4 ANALYSIS OF EXPERIMENTS In order to verify the performance of SMART-RR algorithm based on repetitive protocol factor, this section carries out simulation experiments from three aspects of data traffic, accuracy and privacy protection, and compares SMART-RR algorithm with TAG algorithm and SMART algorithm, selecting building energy consumption and environmental monitoring data including water consumption, electricity consumption, temperature and humidity. We have coded the procedure in MATLAB R2016a and performed all computational experiments on a PC with Intel 2.71GHz CPU and 16GB internal memory. 4.1 Data communication traffic The energy consumption of BEIoT nodes used in building energy consumption data monitoring directly affects the lifetime and availability of sensor networks. From the previous analysis, it can be seen that the energy consumption of BEIoT nodes is mainly consumed in the data communication process of wireless networks. Therefore, the data communication traffic of BEIoT nodes should be reduced as much as possible to effectively control the energy consumption of sensor nodes. Figure 2 Open in new tabDownload slide Data communication traffic of TAG, SMART and SMART-RR. Figure 2 Open in new tabDownload slide Data communication traffic of TAG, SMART and SMART-RR. Figure 3 Open in new tabDownload slide Data communication traffic variation trend with the number of fragments of SMART-RR and SMART. Figure 3 Open in new tabDownload slide Data communication traffic variation trend with the number of fragments of SMART-RR and SMART. Data communication traffic used in BEIoT data transmission and fusion technology mainly includes data transmission and data fusion traffic of node fragmentation. Data fusion traffic is directly related to network size, that is, the data fusion communication traffic is fixed after the BEIoT network structure and the number of nodes is determined. Therefore, it is necessary to compare and analyse the data transmission traffic between SMART-RR algorithm and SMART algorithm, which is also the main traffic overhead of data fragmentation privacy protection algorithm. The data traffic of TAG, SMART and SMART-RR algorithms is shown in Figure 2. Among them, the number of fragments in SMART and SMART-RR is set to three. From the simulation results, we can see that the data communication traffic of SMART and SMART-RR algorithm is less than that of TAG algorithm because of adding data fragmentation strategy, and the data traffic is further compressed because SMART-RR performs some reduction operations on the duplicated data. The relationship between the number of data fragments and the traffic of SMART-RR and SMART algorithm is further analysed. From the simulation results of Figure 3, we can see that with the increase of the number of fragments, the data traffic increases first and then decreases. Figure 4 Open in new tabDownload slide Accuracy variation trend with the number of fragments of SMART-RR and SMART. Figure 4 Open in new tabDownload slide Accuracy variation trend with the number of fragments of SMART-RR and SMART. Figure 5 Open in new tabDownload slide Privacy protection of SMART-RR. Figure 5 Open in new tabDownload slide Privacy protection of SMART-RR. 4.2 Accuracy The SMART-RR algorithm and SMART algorithm proposed in this paper are based on SUM function for data fusion calculation. The data accuracy is defined as: $$ c=\mid{D}_{QS}\mid \kern0.5em /\kern0.5em \mid \sum{D}_i\mid $$ (3) Among them, |$\mid{D}_{QS}\mid$| is the data quantity of the fusion root node and |$\mid \sum{D}_i\mid$| is the original data quantity collected by all sensor nodes. As shown in Figure 4, the data accuracy of SMART-RR and SMART algorithm are sensitive to the number of fragments |$J$|⁠. When |$5\le J\le 7$|⁠, the accuracy tends to increase significantly, and when |$J\ge 8$|⁠, the accuracy tends to be stable. 4.3 Privacy protection The privacy protection of SMART and SMART-RR arithmetic mainly comes from data fragmentation transmission and encryption and decryption mechanism. The data collected by each node is divided into |$J$| slices and sent to the adjacent node |$J-1$| slice data. Only when the eavesdropper obtains all |$J-1$| outgoing links and all inbound links can the complete information sent by the node be decrypted. The probability |$P$| of eavesdropping and decrypting data collected by nodes can be expressed as formula (4): $$ P={p}^{\,J-1}\times{\sum}_{\lambda =0}^D{P}_{d=\lambda}\times{p}^{\lambda } $$ (4) Among them, |$p=k/K$| denotes the probability of data eavesdropping due to key distribution strategy, |$D$| denotes the maximum penetration of all nodes in the network and |${P}_{d=\lambda }$| denotes the probability of |$\lambda$| penetration of nodes. SMART algorithm and SMART-RR algorithm have the same privacy protection as formula (4). TAG algorithm does not introduce privacy protection strategy, so the privacy protection is weaker than the first two algorithms. The privacy protection of SMART-RR algorithm is shown in Figure 5. From the simulation results, it can be seen that SMART-RR algorithm can better protect data privacy by using data fragmentation transmission and encryption and decryption. When |$J\ge 3$|⁠, the probability of eavesdropping on building energy consumption data can be controlled within the range of <0.5%, which can meet the requirements of practical application in the field of building energy consumption and environmental monitoring. The test results also guide us to design a BEIoT-based building energy consumption monitoring system. The number of fragments of node data should be set at more than three slides in order to protect the privacy of building energy consumption data. Based on the above simulation analysis of privacy protection, data traffic and accuracy, it can be seen that the number of data fragments in SMART-RR algorithm directly affects the performance of the algorithm, and more appropriate parameters should be selected according to the specific requirements of privacy protection in practical applications. 5 CONCLUSIONS Aiming at the data transmission and fusion requirements of building energy consumption data acquisition and analysis system, this paper proposes a data fusion algorithm based on repeatability reduction factor, which can compress building energy consumption data with obvious periodicity repeatability, so as to reduce the data communication and energy consumption of sensor nodes and prolong the life cycle of BEIoT. The simulation experiments and comparative analysis verify the performance of the algorithm in data traffic, accuracy and privacy protection. Compared with the same type of algorithm, the algorithm has certain advantages in the field of building energy consumption data acquisition and monitoring applications. Further work can further study the privacy protection algorithms of other data fusion strategies, and collect more practical cases to analyse the application value of the algorithms. FUNDING This work is supported by the project of the National Import Research Priorities Program of China (2016YFB0801004), and the Science and Technology Project of Jiangsu Province Construction System of China (No. 2017ZD225). References [1] Ardakanian O , Bhattacharya A , Culler D . Non-intrusive occupancy monitoring for energy conservation in commercial buildings . Energ Buildings 2018 ; 179 : 311 – 23 . Google Scholar Crossref Search ADS WorldCat [2] Noye S , North R , Fisk D . Smart systems commissioning for energy efficient buildings . Build Serv Eng Res Technol 2016 ; 37 : 194 – 204 . Google Scholar Crossref Search ADS WorldCat [3] Salimi S , Hammad A . Critical review and research roadmap of office building energy management based on occupancy monitoring . Energ Buildings 2019 ; 182 : 214 – 41 . Google Scholar Crossref Search ADS WorldCat [4] Claros L , Padial J , Lauret B . A new and inexpensive open source data acquisition and controller for solar research: application to a water-flow glazing . Renew Energ 2016 ; 92 : 450 – 61 . Google Scholar Crossref Search ADS WorldCat [5] Erkoreka A , Garcia E , Martin K , et al. In-use office building energy characterization through basic monitoring and modelling . Energ Buildings 2016 ; 119 : 256 – 66 . Google Scholar Crossref Search ADS WorldCat [6] Martin A , Millan JA , Bairi A , et al. Environmental monitoring system based on an open source platform and the internet of things for a building energy retrofit . Automat Constr 2018 ; 87 : 201 – 14 . Google Scholar Crossref Search ADS WorldCat [7] Fensel A , Tomic DK , Koller A . Contributing to appliances' energy efficiency with internet of things, smart data and user engagement . Future Gener Comp Sy 2017 ; 76 : 329 – 38 . Google Scholar Crossref Search ADS WorldCat [8] Gupta SP , Pandey UK . Automatic and intelligent integrated system for leakage detection in pipes for water distribution network using internet of things . Smart Innovat Syst Technol 2018 ; 83 : 516 – 23 . Google Scholar Crossref Search ADS WorldCat [9] Jradi M , Arendt K , Sangogboye FC , et al. ObepME: an online building energy performance monitoring and evaluation tool to reduce energy performance gaps . Energ Buildings 2018 ; 166 : 196 – 209 . Google Scholar Crossref Search ADS WorldCat [10] Kalogirou SA . Building integration of solar renewable energy systems towards zero or nearly zero energy buildings . Int J Low-Carbon Technol 2013 ; 10 : 379 – 85 . Google Scholar Crossref Search ADS WorldCat [11] Lord SF , Noye S , Ure JW , et al. Comparative review of building commissioning regulation: a quality perspective . Build Res Inf 2016 ; 44 : 630 – 43 . Google Scholar Crossref Search ADS WorldCat [12] Madden S , Franklin MJ , Hellerstein J , et al. TAG: a Tiny AGgregation service for ad-hoc sensor networks . ACM SIGOPS Oper Syst Rev 2002 ; 36 : 131 – 46 . Google Scholar Crossref Search ADS WorldCat [13] Melendez J , Burgas L , Gamero FI , et al. Fault detection and diagnosis web service module for energy monitoring in buildings . IFAC-PapersOnLine 2018 ; 51 : 15 – 9 . Google Scholar Crossref Search ADS WorldCat [14] Noye S , North R , Fisk D . A wireless sensor network prototype for post-occupancy troubleshooting of building systems . Automat Constr 2018 ; 89 : 225 – 34 . Google Scholar Crossref Search ADS WorldCat [15] Royapoor M , Roskilly T . Building model calibration using energy and environmental data . Energ Buildings 2015 ; 94 : 109 – 20 . Google Scholar Crossref Search ADS WorldCat [16] Xu ZQ , He JL , Li HD . On research IoT-based intelligent parking management system and its design . Int J Smart Home 2016 ; 10 : 217 – 30 . WorldCat © The Author(s) 2019. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
TI - A low power data transfer and fusion algorithm for building energy consumption monitoring
JF - International Journal of Low-Carbon Technologies
DO - 10.1093/ijlct/ctz039
DA - 2019-08-31
UR - https://www.deepdyve.com/lp/oxford-university-press/a-low-power-data-transfer-and-fusion-algorithm-for-building-energy-B6FSUW4n3W
SP - 426
VL - 14
IS - 3
DP - DeepDyve
ER -