Answering your three questions:
1 - Where do I get these signature key?
From Android's own documentation in the section Release Keys
The Android tree includes test-keys under
build/target/product/security
But the next part is where you should really pay attention
Since the test-keys are publicly known, anybody can sign their own
.apk files with the same keys, which may allow them to replace or
hijack system apps built into your OS image. For this reason it is
critical to sign any publicly released or deployed Android OS image
with a special set of release-keys that only you have access to.
So basically unless you can somehow gain access to manufacturer's pvt keys it might be difficult to achieve this. This is why a user in a previous comment was saying this is usually achieved by producing your own build.
2 - Is it going to like a root access If ever I successfully managed
to sign it?
You will not get "root access" by doing it, but you will get access to an extremely high level of access. Specifically, what this achieves you is that you will be granted permissions with declared android:protectionLevel="signature"
which is, arguably, the most exclusive one.
One other dangerous consequence (or fun, depending on how you look at it) of this is that you can now run your app under system user process android:sharedUserId="android.uid.system"
- under android's "process sandboxed" security rules this would normally fail.
3 - What is the difference between Root vs Signed with key?
With an app signed with the platform key from your build, you can get the permissions mentioned above, or run your app with UID 1000 (system uid) which in android is much more powerful than the UIDs of other apps because of the permissions it can request, this is a behaviour specific of Android though.
In a rooted device, you can use UID 0 (root) which has the broadest access in linux based systems, you can bypass most of the security sandboxing/checks/fences on the OS.
Hope this helps ;)