Search

With the rapid growth of renewable energy generation, curtailment condition of wind and photovoltaic power has grown significantly and become a serious problem. In this context, this paper proposes a new method to evaluate renewable power accommodation capability based on typical scenario construction. Improved FCM clustering algorithm is used to screen raw photovoltaic and wind power generation data and construct different generation scene to improving the accuracy and efficiency of computation. The renewable energy accommodation capabilities of different typical scenarios were assessed in this paper. Taking province in east China as an example, renewable power accommodation capabilities of typical scenarios in 2018 are assessed. The simulation results show that the optimization method based on typical scenario construction can promote the efficiency and validity of renewable energy accommodation capability assessment.

High write throughput plays a vital role in improving the time efficiency of RFID-enabled applications, such as password update and over-the-air programming. In this paper, we make a fresh attempt to study the under-investigated problem of group write and discuss how to improve the write throughput in the commodity RFID system. By conducting extensive experiments, we reveal that the select operation specified by the C1G2 standard takes up the large overhead of a write cycle, which is the key factor in determining the write efficiency. Based on this finding, we propose an efficient write bundling (WB) scheme that bundles multiple writes up and executes them together in a burst mode, which greatly reduces the number of selects and thereby amplifies the write throughput. In WB, a carefully designed bit vector is assigned to each tag for connecting multiple tags into a run, such that a single select command is able to pick these tags concurrently. Besides, WB integrates multiple select commands into one by running a series of logic operations on the inventoried flags, which are used to indicate whether or not a tag is active. We implement WB in a commodity RFID system, with no need of any software modifications or hardware argument. Extensive experiment results show that WB is able to reduce the number of selects by 87.5% and produce a $2\times$ write amplification, compared with the C1G2-compatible exclusive write.

In RFID systems, a cloning attack is to fabricate one or more replicas of a genuine tag, so that these replicas behave exactly the same as the genuine tag and fool the reader for getting legal authorization, leading to potential financial loss or reputation damage for the corporations. These replicas are called clone tags. Although many advanced solutions have been proposed to combat cloning attack, they need to either modify the MAC- layer protocols or increase extra hardware resources, which cannot be deployed on commercial off-the-shelf (COTS) RFID devices for practical use. In this paper, we take a fresh attempt to counterattack tag cloning based on COTS RFID devices and the universal C1G2 standard, without any software redesign or hardware augment needed. The basic idea is to use the RF signal profile to characterize each tag. Since these physical-layer data are measured by the reader and susceptible to various environmental factors, they are hard to be estimated by the attackers; let alone be cloned. Even so, we assert that it is challenging to identify clone tags as the signal data from a genuine tag and its clones are all mixed together. Besides, the tag moving has a great impact on the measured RF signals. To overcome these challenges, we propose a clustering-based scheme that detects the cloning attack in the still scene and a chain- based scheme for clone detection in the dynamic scene, respectively. Extensive experiments on COTS RFID devices demonstrate that the detection accuracy of our approaches reaches 99.8% in a still case and 99.3% in a dynamic scene.

In our daily life, some sensitive cargos (e.g., refrigerators) are required to keep one right side up. If these items are turned upside down due to incorrect transportation or wrong storage, some unpredictable exception will happen, leading to potential economic loss. In this paper, we propose a lightweight system TagUP that uses RFID to detect the upside-down exception. By dynamically changing the frequency of the electromagnetic wave and comparing the phase difference between two measurements, TagUP is able to achieve the detection accuracy of 98.0% when two tags are placed 60cm apart.