For the 1st part,
Regular languages are closed under Reversal ($L^R$) and Concatenation ($L.L'$) operations.
Formally,
$L$ is Regular $\iff$ $L^R$ is Regular
$L$ and $L'$ are Regular $\implies$ $L.L'$ is Regular
So, $L$ and $L^R$ are Regular $\implies$ $L.L^R$ is Regular
(You can go through the proofs for these 2 properties from any standard book, but I don't think it is required to write for this question)
For 2nd part,
Steps to Contruct FA $N'$ for $L^R$ given a FA $N$ for $L$: (from my answer for 2011-B-04a)
- Convert the NFA $N$ to a DFA using the subset construction algorithm. Let the resulting DFA be $D$.
- To build an NFA $N'$ to accept $L^R$, we do the following:
- Reverse the direction of all edges in the transition diagram.
That is, replace every transition $\delta(q_i, a) = q_j$, where $a \in \{0,1\}$, with $\delta^R(q_j, a) = q_i$. - The original start state of $D$ becomes the new final state in $N'$.
- The final state(s) in $D$ become the new start state(s) in $N'$. If $|F| > 1$, add a new start state $q_0$ and add $\epsilon$-transitions from $q_0$ to all original final states in $D$.
Now we have finite automata for $L$ and $L^R$, which are $N$ and $N'$ respectively.
We make a new FA say $M$ for accepting $L.L^R$, which has following properties:
- M has all the states from $N$ and $N'$. If $i$ and $j$ are number of the states in $N$ and $N'$, then $M$ will have $i+j$ states.
- The start state of $M$ is the original start state ($q_0$) of $N$.
- The only final state(s) of $M$ are the original final state(s) of $N'$.
- We add $\epsilon$-Transitions from every state in $M$ which were originally final state(s) in $N$ to the state which was originally start state in $N'$.
So what really happens in this new FA $M$?
I will try to to give an informal proof of working for $M$.
For a string satisfying the language $L.L^R$ we need to show that it reaches one of the final state of $M$.
First the string moves to a state which was a final state in $N$, satisfying the condition for language $L$.
Then via a $\epsilon$-transition,it reaches a state in $M$ which was a start state in $N'$, from there satifying the condition for language $L^R$, the string reaches one of final states in $M$, as they were also final states in $N'$.
I hope this answer was helpful!