Due to better efficiency, gain and thermal performance compared to other semiconductor technologies, GaN power amplifiers are very attractive in the present 5G era. Meanwhile, accurate GaN HEMT device modelling is one of the critical steps to design PAs successfully. Therefore, research on high-frequency GaN HEMT device modelling method is of vital importance. This paper first presents a wideband GaN HEMTs model for 5G power amplifiers. A loadpull system available for 10-67 GHz measurement is set up to obtain wideband S parameter and RF performance results. The whole GaN device modelling could be divided into two parts: the small signal modelling and the large signal modelling. Direct optimization method with polynomial fitting is employed to obtain equivalent small signal circuit parameters, which improve the accuracy and efficiency of parameter extraction. Also, artificial neural network (ANN) technique is utilized to build charge and nonlinear current models, which takes the self-heating and trapping effects into consideration in the large signal modelling. The ANN technique could substitute the complex empirical equations as other papers has reported, and thus makes the extracted parameters less and the extraction process more accurate and efficient. At last, the proposed model is implemented and verified in ADS, the error between the measurement and simulation results is less than 5%.

Recently, there is an increasing demand for better in-band(IB) and out-of-band(OOB) linearity for sub-6GHz applications [1 – 3]. To achieve high linearity, one solution is to utilize current-mode direct conversion [1]. In this architecture, the low-noise amplifier (LNA) acts as a transconductor, which avoids large voltage gain, thus achieving high linearity. Another solution is to employ mixer-first architecture [2,3]. This kind of receiver starts with a passive mixer, which avoids any possible voltage gain brought by the LNA and shows improved tolerance to the blockers. However, unlike the current-mode architecture, its transimpedance-amplifier (TIA) input is not a virtual ground, but a high-impedance node, for input matching purpose [2]. Meanwhile, to achieve better NF, noise-cancellation techniques are widely used in different architectures to improve the noise performance [4,5], achieving sub-1dB NF at 1GHz [4] and 1.75dB NF at 5GHz [5]. However, both receivers are with LNA-first architecture, which limits their linearity.

Bitcoin security draws more and more attention recently. One of Bitcoin vulnerabilities is caused by ECDSA weak randomness. A random number is not cryptographically secure, which leads to private key leakage and even fund theft. This security problem has been well known in Bitcoin community and fixed by applying RFC 6979 update in 2013. In this paper, we systematically revisit the cases where random numbers are reused and evaluate them based on practical Bitcoin transactions. After analyzing Bitcoin transaction dataset from January 2009 to July 2017, we find that there are still approximately 0.48 percent of transactions involving this vulnerability, and 1331 private keys have been compromised. In addition, the transactions related to some involved addresses have a common pattern, which gives us a clue that a spam transaction attack may take advantage of ECDSA weak randomness. We also examine mainstream Bitcoin software wallets to check whether they are susceptible to ECDSA weak randomness. Even the result is quite optimistic, an example that one of the influenced addresses leaked in April 2014 is still in use again in August 2017 reflects that the severity of ECDSA weak randomness may not be paid enough attention even after its discovery and solution in 2013. (C) 2019 Elsevier B.V. All rights reserved.