At Eurocrypt 2005, Brent Waters proposed an efficient Identity Based Encryption scheme which is secure in the standard model. One drawback of this scheme is that the number of elements in the public parameter is rather large. Here we propose a generalisation of Waters scheme. In particular, we show that there is an interesting trade-off between the tightness of the security reduction and smallness of the public parameter. For a given security level, this implies that if one reduces the number of elements in public parameter then there is a corresponding increase in the computational cost due to the increase in group size. This introduces a flexibility in choosing the public parameter size without compromising in security. In concrete terms, to achieve 80-bit security for 160-bit identities we show that compared to Waters protocol the public parameter size can be reduced by almost 90 % while increasing the computation cost by 30%. Our construction is proven secure in the standard model without random oracles. Additionally, we show that CCA security can also be achieved through the reduction to oracle decision bilinear Diffie-Hellman problem (OBDH).