If you want to avoid any type of hardware engineering your options are somewhat limited.

The type of sources you can get entropy from include:

1) content and timing of IP packets the computer sees

2) keyboard and mouse data (and timing)

3) microphone

4) webcam

5) clock

The problem with some (or most) of these sources of entropy is that they are somewhat predictable (and biased), to some degree. This means that you should probably have to post-process the random data so that you get output sequences with good properties (uniform, uncorrelated and unpredictable).

Some examples you should look into, where people try to do this (i.e. generate good quality random numbers from medium quality entropy sources), include the "Fortuna PRNG" and "Lavarand" (just look them up on Wikipedia). In a practical sense, both these systems are based on using random data to seed a pseudo-random number generator.

But, yeah, unless you are willing to actually go into hardware engineering, it is not easy to ensure you have a good reliable source of entropy for your RNG (particularly if you require random numbers at boot time).

True randomness can only be gained from natural sources. Examples could be the pattern of radio signals, noise around you or at some specific place. You will need hardware (sensors) for that. If we use hardware for generation we can get Pseudo-Random Numbers, which seem to be true random numbers but they are not. Cellular Automata (CA) is an example of Pseudo-Random Number Generators (PRNG).   

What Muhammad says is true. Luckily, most computers have a few hardware sensors built in (e.g. microphone and webcam).

For completeness, here are some other sources of entropy: https://en.wikipedia.org/wiki/CryptGenRandom#Method_of_operation

There are two main methods used to generate random numbers.

1. The method measures some physical phenomenon that is expected to be random and then rewards for possible biases in the measurement process. Example sources include measuring atmospheric noise, thermal noise, and other external electromagnetic and quantum phenomena. For example, cosmic background radiation or radioactive decay as measured over short timescales represents sources of natural entropy. The speed at which entropy can be produced from natural sources is dependent on the underlying physical phenomena being measured. Therefore, sources of naturally occurring 'true' entropy are said to be blocking i.e. rate-limited until enough entropy is produced to meet demand. On some Unix-like systems, including Linux Distributions and BSD Branches, the pseudo device file /dev/random will block until sufficient entropy is produced from the environment. Due to this blocking behavior large bulk reads from/dev/random, such as filling a hard disk with random bits, can often be slow.

2. The second method uses computational algorithms that can produce long sequences of apparently random results, which are in fact completely determined by a shorter initial value, known as a seed or key. The latter type is often called pseudorandom number generators. These types of generators do not typically rely on sources of naturally occurring entropy, though they may be periodically seeded by natural sources, they are non-blocking i.e. not rate-limited by an external event.

A "random number generator" based solely on deterministic computation cannot be regarded as a "true" random number generator in the purest sense of the word, since their output is inherently predictable if all seed values are known. In practice however they are sufficient for most tasks. Carefully designed and implemented pseudo-random number generators can even be certified for security-critical cryptographic purposes, as is the case with the yarrow algorithm and fortuna (PRNG). The former being the basis of the/dev/random source of entropy on FreeBSD, AIX, Mac OS X, NetBSD and others. OpenBSD also uses a pseudo-random number algorithm based on RC4 known asarc4random.

Jaroslav,

Listen to me :) Use quantum. Go to the nearest quantum optics lab and borrow a photon counter. Put some CW light onto it (from LED for example) so that it counts about 1 MHz and you got yourself a random event generator. The simplest way to go from there to random numbers is described here: https://www.researchgate.net/publication/234953799_Fast_nondeterministic_random_bit_generator_based_on_weakly_correlated_physical_events?ev=prf_pub

Article Fast nondeterministic random bit generator based on weakly c...

P.S. If you do not want to engage in generating true random numbers yourself there are quantum random number generators connected to the web from where you can download superb quality randomness, like for example: https://qrng.physik.hu-berlin.de/

Check out Random.org

its an awesome resource! I used it in my freelance project 2 years back.

I agree to majority of comments shared here

Good luck.

This is an example of what may solve your need http://www.random.org/integers/ In fact you can use any source of random phenomena which can be converted to an electrical signal, then digitized by an A2D converter and use the output values as random numbers. 

As a paraphrase, nothing in the Nature is random. Even what we call "chaos" has its deterministic rules. The truth is that the human kind is still far away from really understanding the rules of Nature. So we, the humans, "blame" Nature for so many phenomena which we are unable yet to understand in full their rules (human cognitive dissonance).  

Within electronic, digital computational environments, there is no such thing as a random number. The best one can do is to distribute probability of occurrence across a set of options, and the number of those options is related to the number of bits that are taken to represent any one such option. Hence, the proper understanding of randomness within computational environments is in terms of disorder; this is the best you can do, to obtain maximal disorder; and thus, predictability in a maximal sense.

I think this work can provide you guiadence.

https://link.springer.com/article/10.1007%2Fs11071-019-05112-4

https://rdcu.be/bKhar