The study of continuum sky background spectrum at low radio frequencies has achieved specific importance in present day observational cosmology . At these low frequencies the sky continuum is contributed by the extragalactic radio sources together with the synchrotron emission of the Milky Way as well as CMB. Following the recombination, the energy exchange between the primordial neutral hydrogen and CMB photons, during its propagation through the ”Dark ages” as well as the ”cosmic dawn” resulted in absorption and emission features in CMB spectrum which evolved with the evolution of the HI over cosmic time. Due to cosmological expansion of the Universe such spectral signatures of cosmological origin is now redshifted to low radio wavelengths. Although the peak to peak amplitudes of the same are smaller by orders of magnitude than the total galactic and extragalactic contribution at these frequencies the later is expected to be smooth over the scales of few hundred MHz. Hence, except for the extreme cases where the time scale over which such radiative transfer interaction occurred is very long and therefore the spectral fluctuations are spread over a large range of frequencies, these cosmological radiation signatures should be detectable at meter wavelengths. The duration and frequency at which such spectral signatures may occur can give constraints on the physical processes that governed the process of such energy exchange at a very early time, the history of evolution of the gas and the nature and evolution of sources of first light in the Universe. Measurements of the absolute brightness of the continuum background at meter wavelengths and detection of the spatial and spectral variations can therefore be an important probe of cosmology. In addition, measurements to date suggest that the radio background that is of extragalactic origin consists of CMB plus a power-law spectrum and has a brightness temperature of (1.2 ± 0.09) × (ν/1 GHz)−2.60±0.04 K. Surprisingly, the sky brightness corresponding to discrete radio sources detected in the deepest surveys to date account for only a fraction of the extragalactic radio background, even after excluding the CMB. Improved measurements of the radio background and, in particular, the spectrum at long wavelengths where errors are relatively larger, are important in estimating the spectrum of the unexplained part and thereby constraining the sources of this cosmic radiation. The wideband measurements at meter wavelengths pose limit on the accuracy of such measurements where the errors are relatively large. The instrument systematics, which are frequency dependent, in interaction with the sky signal may result in an incorrect estimate of the absolute sky brightness as well as may give rise to spectral features which may confuse with the true cosmological signatures in the foreground. A strategic system design with the aim of minimizing the systematics and characterization of the system non-ideal behavior can lead to the measurement accuracy with which the cosmological signatures could be detected along with the absolute measurement of the foreground. The aim of the work that constitute this thesis is precision measurement of the continuum radio emission at long wavelengths and detect the signature from the epoch of reionization in the background spectrum. A single element radio telescope system has been designed and built which is capable of useful measurement of spectral signatures of the EoR in the radio background. SARAS deployed in the Gauribidanur Observatory, about 80 km north of Bangalore in India. The design, calibration method and observation strategies developed are novel and unique, and relevant for any wideband measurements. The content of this thesis is outlined below. The Chapter 1 briefly introduces to the 21 cm cosmology. The neutral hydrogen as an cosmological probe is discussed first. The redshifted 21 cm signal and its possible use to probe the early Universe is discussed in detail. The chapter also gives a description of the redshifted 21 cm background and the EoR global signature as well as the 21 cm power spectrum. Finally an account of the recent and future experiments for detection of the global EoR signature and the EoR power spectrum are given. In the light of the science introduced in this chapter, the major aim and the work of the thesis is also summarised. In Chapter 2 the design philosophy of the SARAS spectrometer is illustrated. The basic elements of the system, the antenna, analog and the digital receiver have been described in detail. The complete configuration and the complex performance of the integrated system are detailed. For an ideal performance of such a wideband system the calibration strategies are considered and measurement equations are derived. Chapter3 addresses the most important issues of systematic effect for this wideband continuum measurement. The designed system performance deviates from its ideal due to non-idealities. The non-ideal behaviour of the real systems that are the limiting factors for a precision wideband measurement at low frequencies are thoroughly investigated. The effects of single and multiple reflections that occurs due to impedance mismatch internal to the system and results in spurious response in the measured data are considered. Within the limits of the known systematics, the system has been argued to qualify to be able to measure the EoR spectral signature in the background spectrum. Chapter 4 gives the details of the SARAS observation and description of the measurement that has been done by the SARAS spectrometer from the Gouribidanur Radio Observatory . The detailed understanding of the systematics led to the two strategic observation modes; EoR mode and RB mode, which are also discussed in this chapter. Next, the absolute calibration of the SARAS spectrometer is described following which the interference rejection algorithms, purpose developed for rejection of RFI from the SARAS data are outlined. The measured data is calibrated and the additional calibration product is derived. The data set is then modelled by modelling the sky noise and the systematic effects. In Chapter 5 the methodology of the SARAS data analysis is illustrated along with step by step analysis of the observation made in the RB mode. The model that is derived in chapter 4 is fitted to the measurements and the parameters are estimated by minimising the merit function Chi-square. The minimisation followed the downhill simplex algorithm which is outlined in this chapter. The data analysis strategy relied on the derivation of the initial values of the parameters from the measured data. The data has been fitted in hierarchy and initial guesses for subset of parameters are derived from each step of hierarchical modelling. This method of analysis is strategic and discussed in detail in this chapter. The statistical and systematic error of measurement are discussed next. Finally, the posterior probability distribution of the parameters are calculated by uniformly sampling the parameter space around the best fit values and calculating the Chi-square and the likelihood functions of the parameters. Mariginalizing the computed posterior probability distribution over the system parameters, the error on estimated sky parameter or the confidence region of the sky parameter is estimated. Chapter 6 presents the analysis identical to that presented in the Chapter 5 for the data acquired in the second observing mode, the EoR mode is presented. In this section the detection ability and the limitations of the SARAS observations, made from the Gouribidanur Radio observatory, for a useful detection of the EoR spectral signature is considered. Chapter 7 described a concept of system bandpass calibration using the time domain information that could be obtained from the voltage samples before computing the power spectrum. In the Chapter 3, the spurious periodic correlation, generated due to the internal reflections of the noise voltages is discussed in detail. If a short pulse is injected into the system via the antenna, due to internal reflections, the primary pulse in the output voltage samples would be followed by a series of delayed pulses each of which is generated by subsequent reflections internal to the system. Inspection of such pulses in the time domain could potentially provide with accurate bandpass calibration. A digital hardware is programmed and a nano second pulse generator-accumulator system is built. The performance of this system and table top demonstration of this bandpass calibration concept is presented in this chapter. This pulse injection system is used in a proposed calibration experiment in the Parkes radio telescope. The initial result of the experiment is presented in this chapter. In Chapter 8 the work carried out for this thesis is summarised. The end to end task of carrying out the background measurement by purpose building a single element radio telescope, observing with it and analysis of the data has led to a successful measurement of the background spectrum with an accuracy > 1%. The detailed understanding of the problems associated with the precision measurements and development of the measurement techniques that can overcome such problems has led to significant progress towards a successful detection of the EoR signatures. Some aspects and ideas which are understood to be essentials for such an experiment yet unexplored here due to limited time are listed in this section and the future prospcts of this work is also discussed.

No description available.