-
Towards Machine Learning on data from Professional Cyclists
Authors:
Agrin Hilmkil,
Oscar Ivarsson,
Moa Johansson,
Dan Kuylenstierna,
Teun van Erp
Abstract:
Professional sports are developing towards increasingly scientific training methods with increasing amounts of data being collected from laboratory tests, training sessions and competitions. In cycling, it is standard to equip bicycles with small computers recording data from sensors such as power-meters, in addition to heart-rate, speed, altitude etc. Recently, machine learning techniques have pr…
▽ More
Professional sports are developing towards increasingly scientific training methods with increasing amounts of data being collected from laboratory tests, training sessions and competitions. In cycling, it is standard to equip bicycles with small computers recording data from sensors such as power-meters, in addition to heart-rate, speed, altitude etc. Recently, machine learning techniques have provided huge success in a wide variety of areas where large amounts of data (big data) is available. In this paper, we perform a pilot experiment on machine learning to model physical response in elite cyclists. As a first experiment, we show that it is possible to train a LSTM machine learning algorithm to predict the heart-rate response of a cyclist during a training session. This work is a promising first step towards developing more elaborate models based on big data and machine learning to capture performance aspects of athletes.
△ Less
Submitted 1 August, 2018;
originally announced August 2018.
-
Oscillator Phase Noise and Small-Scale Channel Fading in Higher Frequency Bands
Authors:
M. Reza Khanzadi,
Rajet Krishnan,
Dan Kuylenstierna,
Thomas Eriksson
Abstract:
This paper investigates the effect of oscillator phase noise and channel variations due to fading on the performance of communication systems at frequency bands higher than 10GHz. Phase noise and channel models are reviewed and technology-dependent bounds on the phase noise quality of radio oscillators are presented. Our study shows that, in general, both channel variations and phase noise can hav…
▽ More
This paper investigates the effect of oscillator phase noise and channel variations due to fading on the performance of communication systems at frequency bands higher than 10GHz. Phase noise and channel models are reviewed and technology-dependent bounds on the phase noise quality of radio oscillators are presented. Our study shows that, in general, both channel variations and phase noise can have severe effects on the system performance at high frequencies. Importantly, their relative severity depends on the application scenario and system parameters such as center frequency and bandwidth. Channel variations are seen to be more severe than phase noise when the relative velocity between the transmitter and receiver is high. On the other hand, performance degradation due to phase noise can be more severe when the center frequency is increased and the bandwidth is kept a constant, or when oscillators based on low power CMOS technology are used, as opposed to high power GaN HEMT based oscillators.
△ Less
Submitted 25 September, 2014; v1 submitted 16 July, 2014;
originally announced July 2014.
-
Calculation of the Performance of Communication Systems from Measured Oscillator Phase Noise
Authors:
M. Reza Khanzadi,
Dan Kuylenstierna,
Ashkan Panahi,
Thomas Eriksson,
Herbert Zirath
Abstract:
Oscillator phase noise (PN) is one of the major problems that affect the performance of communication systems. In this paper, a direct connection between oscillator measurements, in terms of measured single-side band PN spectrum, and the optimal communication system performance, in terms of the resulting error vector magnitude (EVM) due to PN, is mathematically derived and analyzed. First, a stati…
▽ More
Oscillator phase noise (PN) is one of the major problems that affect the performance of communication systems. In this paper, a direct connection between oscillator measurements, in terms of measured single-side band PN spectrum, and the optimal communication system performance, in terms of the resulting error vector magnitude (EVM) due to PN, is mathematically derived and analyzed. First, a statistical model of the PN, considering the effect of white and colored noise sources, is derived. Then, we utilize this model to derive the modified Bayesian Cramer-Rao bound on PN estimation, and use it to find an EVM bound for the system performance. Based on our analysis, it is found that the influence from different noise regions strongly depends on the communication bandwidth, i.e., the symbol rate. For high symbol rate communication systems, cumulative PN that appears near carrier is of relatively low importance compared to the white PN far from carrier. Our results also show that 1/f^3 noise is more predictable compared to 1/f^2 noise and in a fair comparison it affects the performance less.
△ Less
Submitted 10 December, 2013; v1 submitted 6 May, 2013;
originally announced May 2013.
