NAT is network access translation. You have 1 public IP and several private LAN IP addresses. When a LAN pc requests info that is provided by the Internet, the router translates the LAN IP to the router public IP so it can be routed over the Internet and a reply can be provided back. This also is a firewal function so that inbound data packets that are not a reply to a specific request are discarded.
The / 28 means a 28 bit subnet mask (255.255.255.240). In a 192.168.18.x series, the subnets are 192.168.18.16, 192.168.18.32, 192.168.18.48, 192.168.18.64, 192.168.18.80, 192.168.18.96 etc. Note that there is a difference of 16 in the last set of digits. This identifies the subnet and neither the subnet ID or the max number for the subnet can be used for routing as they are reserved for subnet ID and for broadcast respectively.
Therefore if you are using the 192.168.18.48 / 28 subnet
1. address = 192.168.18.48 and not used as it is an identifier
2. address = 192.168.18.63 is for broadcast and not used as an identifier.
3. Of the 16 IP addresses only 14 can be used as an identifier. One of these must be the default gateway IP, leaving 13 other for LAN use.
4. The default gateway is the IP address of the device to which packets are forwarded when sending devices do not know where to route them. In very small networks this is often a router's LAN IP address; the router's WAN address is the public IP address.
5. I have given you the LAN private IP addresses for each pc to connect as shown above. The default gateway you specified. If you are giving out static IP addresses, use the default gateway address you list above, subnet 255.255.255.240 (28 bit) and make sure you do not use the .40, .56, and .63 addresses as indicated above. Use the default gateway as the DNS.
There are definite calculation methods or formulas. This is rather elaborate and lengthy and you are best served by getting a text book and learning from there as it constitutes a few chapters.